Kiev/Moscow 1941: The Gornostaipol Option

[per70]

It is common knowledge that the OstHeer suffered severe logistical difficulties in late October through the winter of 1941/1942. I shouldn't have to provide detailed explanations of basic facts that are known to every regular user on this forum. The only thing in your ATL that could change this would be an earlier capture of Moscow, allowing German rail lines to be extended to that city by early December (possibly).

A brief summary of (my impression) of the logistical problems the Germans experienced during the IV quarter of 41:

1. They didn't attack along the main highway, instead opting to attack in a weaker defended sector with weaker infrastructure.
2. In the first ten days of the campaign, they had to manage a severe road congestion problem, as the transport columns had to fight for road space with infantry forces advancing forward.
3. The key to solving the logistical challenge was fixing the railroads, but in order to do that, they had to fix the railroad bridges.
a) key nodes had been bypassed initially. For instance, the railroad bridge at Briansk wasn't secured until the fourth week of the campaign.
b) and once secured, they needed several weeks of repair time to get fully operational
4. While the rail lines were out of order, they were dependant on trucks, which had to drive on often damaged, muddy secondary or tertiary roads, which caused tremendous difficulties.
5. By mid- to late- November, the rail lines started getting operational, although they had only focused on a select few lines. And there had been inadequate time to upgrade the lines/associated buildings, making the entire network fragile.
6. Furthermore, with the troops desperate for supplies - having been shortchanged for close to two months - the Germans gambled that the winter would be delayed for a few more weeks, and gave higher priority to supplies over railroad equipment.
7. When the gamble failed - and the weather did turn sour - the fragile railroad network collapsed. With each failure creating negative ripple effects and so forth.

If correct, some of those problems would be diminished with a Sep 1 attack.
1, 2 and 3 would be the same.
But 4 would be easier in the period Sep 10 - Oct 10, as compared to Oct 10 - Nov 10.
And just as the mud season hit in October, the railroad lines - which were less vulnerable to the mud - would come online.
Making the mud less of a problem supply wise - although it would still restrict combat operations.
And finally, by early December, the Germans would have spent about six weeks (as opposed to two) making the network more robust.

The downside to all this would be a less developed rail network in the Smolensk area, as the focus would be shifted east.

[TheMarcksPlan]

5. By mid- to late- November, the rail lines started getting operational, although they had only focused on a select few lines. And there had been inadequate time to upgrade the lines/associated buildings, making the entire network fragile.
6. Furthermore, with the troops desperate for supplies - having been shortchanged for close to two months - the Germans gambled that the winter would be delayed for a few more weeks, and gave higher priority to supplies over railroad equipment.

Interesting. Source? I'm digging through the archives for data on the railways.

the railroad bridge at Briansk wasn't secured until the fourth week of the campaign.

At least relative to campaign start, it would be delayed further in this ATL, as it's a two-PzGr attack lacking the southern element initially.

And just as the mud season hit in October, the railroad lines - which were less vulnerable to the mud - would come online.
Making the mud less of a problem supply wise - although it would still restrict combat operations.
And finally, by early December, the Germans would have spent about six weeks (as opposed to two) making the network more robust.

Makes sense. The timing was almost optimally bad for Germany OTL: extend the lines in the last bit of good weather, forcing the needed rail work period to extend into the beginning of freezing. The archives are replete with mentions of getting rail work done before the winter.

It is common knowledge that the OstHeer suffered severe logistical difficulties in late October through the winter of 1941/1942. I shouldn't have to provide detailed explanations of basic facts

The above discussion with per70 is of the kind that is necessary. Calling things good or bad and leaving it at that is uninteresting, unhelpful.

[historygeek2021]

The above discussion with per70 is of the kind that is necessary. Calling things good or bad and leaving it at that is uninteresting, unhelpful.

The burden is on you to show how your ATL would improve the logistical situation of the OstHeer to be able to accomplish the "kessels" that would enable the "delta" of 2 milllion more Red Army permanent casualties over the OTL. Flipping that burden onto me by feigning ignorance of a basic fact (the OstHeer's logistics were insufficient in the winter of 41/42) is precisely what is unhelpful, and your constant resort to insults of other people on this forum (calling their writing "uninteresting") is poisoning what would otherwise be a very interesting and informative discussion.

[History Learner]

The above discussion with per70 is of the kind that is necessary. Calling things good or bad and leaving it at that is uninteresting, unhelpful.

The burden is on you to show how your ATL would improve the logistical situation of the OstHeer to be able to accomplish the "kessels" that would enable the "delta" of 2 milllion more Red Army permanent casualties over the OTL. Flipping that burden onto me by feigning ignorance of a basic fact (the OstHeer's logistics were insufficient in the winter of 41/42) is precisely what is unhelpful, and your constant resort to insults of other people on this forum (calling their writing "uninteresting") is poisoning what would otherwise be a very interesting and informative discussion.

And it's a burden he's already responded to you over, and in great detail:

Maybe. Let me explain the whole context, then I'll get back to your other points later.

2mil is an upside case. It's based on the following conditions:

• September delta to PoW: ~500k
• October being slightly higher than OTL PoW in AGC sector, based on a large Kessel at Moscow. 650k there would keep us even with OTL October's Vyazma.
• In AGS's sector, the "additional panzer group" adding another delta in a pocket such as the Kharkov Kessel sketched in the ATL. 200k delta?
• In AGN's sector: Leningrad and Volkhov fronts suffer less damage in September (no XXXIX MC) but what about October-November? If the foregoing disasters have befallen RKKA, those fronts are surely weaker with ~700k more prisoners taken front-wide. RKKA's front strength in OTL November as 2.2mil: ATL it's ~30% weaker. In that condition, I can't see Volkhov Front stopping AGN from reaching the Svir (especially if all PzGr4 stays with it), which would mean losing Volhov and Leningrad fronts, with the latter starving out at some point that winter. Together that's at least another 700k permanent losses. Now we're at 1.4mil more permanent losses.
• Then there's the prospect of slow German pushes during the winter - think extending the November advances, probably after a rest/refit interlude (so February/March). During OTL November, Ostheer was capturing ~3 PoW daily for every 1,000 RKKA front strength. Once Ostheer was on the back foot, a few % of that. An Ostheer "slow push" winter offensive against ~3mil RKKA in January/February would, on OTL November trends, bag >500k PoW/day absent any Kessels. Even prior to November, the rate was ~2/day per 1,000 RKKA. I have not seen the literature properly account for the abnormal rate of Soviet "tactical" surrenders when Ostheer was advancing but the evidence is indisputable.

Adding up all the foregoing gives something like 2mil more PoW - assuming L'grad and Moscow fall.

This a non-linear process: Winning more in September makes winning more in October more likely, which makes winning in November...

Likewise, your continued assertions about the logistics of AGC in this situation have been responded to by Per70, who you have not responded to quite transparently. Your passive aggressive style of posting is neither informative or substantive, but it also does nothing to advance the conversation other than to serve as a pretty clear tactic to not actually have to engage in rebuttals that pick apart your specific criticisms.

[historygeek2021]

And it's a burden he's already responded to you over, and in great detail:

Maybe. Let me explain the whole context, then I'll get back to your other points later.

2mil is an upside case. It's based on the following conditions:

• September delta to PoW: ~500k
• October being slightly higher than OTL PoW in AGC sector, based on a large Kessel at Moscow. 650k there would keep us even with OTL October's Vyazma.
• In AGS's sector, the "additional panzer group" adding another delta in a pocket such as the Kharkov Kessel sketched in the ATL. 200k delta?
• In AGN's sector: Leningrad and Volkhov fronts suffer less damage in September (no XXXIX MC) but what about October-November? If the foregoing disasters have befallen RKKA, those fronts are surely weaker with ~700k more prisoners taken front-wide. RKKA's front strength in OTL November as 2.2mil: ATL it's ~30% weaker. In that condition, I can't see Volkhov Front stopping AGN from reaching the Svir (especially if all PzGr4 stays with it), which would mean losing Volhov and Leningrad fronts, with the latter starving out at some point that winter. Together that's at least another 700k permanent losses. Now we're at 1.4mil more permanent losses.
• Then there's the prospect of slow German pushes during the winter - think extending the November advances, probably after a rest/refit interlude (so February/March). During OTL November, Ostheer was capturing ~3 PoW daily for every 1,000 RKKA front strength. Once Ostheer was on the back foot, a few % of that. An Ostheer "slow push" winter offensive against ~3mil RKKA in January/February would, on OTL November trends, bag >500k PoW/day absent any Kessels. Even prior to November, the rate was ~2/day per 1,000 RKKA. I have not seen the literature properly account for the abnormal rate of Soviet "tactical" surrenders when Ostheer was advancing but the evidence is indisputable.

Adding up all the foregoing gives something like 2mil more PoW - assuming L'grad and Moscow fall.

This a non-linear process: Winning more in September makes winning more in October more likely, which makes winning in November...

No, nothing in the text you quoted addresses the issue of logistics.

Likewise, your continued assertions about the logistics of AGC in this situation have been responded to by Per70, who you have not responded to quite transparently. Your passive aggressive style of posting is neither informative or substantive, but it also does nothing to advance the conversation other than to serve as a pretty clear tactic to not actually have to engage in rebuttals that pick apart your specific criticisms.

Piling on insults against me because I disagree with you isn't helpful to the discourse of the thread. Cut it out.

[per70]

Let's try to calm down, and continue without the insults. Please.

[TheMarcksPlan]

The burden is on you to show how your ATL would improve the logistical situation of the OstHeer

Actually it perhaps hasn't been specified adequately who has the burden of proof (differentiable into burden of persuasion and burden of production) here. The concept of assigning burden has good applicability.

Even before getting into burden analysis, we have to specify the relevant "proceedings" in which burden is being assigned. I.e. at what stage of ATL proceedings do you see the logistical issues becoming dispositive?

The Gornostaipol prong of the Kiev offensive? Seems unlikely that's your beef, so I'll skip that for now.

A September 1 Vyazma? In the other thread most conceded this would be logistically feasible, can't remember whether you did. Skipping that one for now as well.

The post-Vyazma lunge at Moscow? That seems the most likely stage of proceedings. My argument is that the railheads would advance to ~Vyazma during September and when the Moscow lunge occurs AGC's logistics will be roughly as good as for OTL Vyazma. I.e. it would be receiving roughly as many trains/day when setting out for Moscow as it did when setting out for Vyazma OTL. Perhaps a bit less good. On the flip side, AGC is facing a weaker October opponent because in ATL October Vyazma and Kiev have already occurred (plus some AGC flank-cleaning during the remainder of September - think Roslavl-style medium kessels farther east).

As logistics is not an either/or but a continuum that influences combat power (up to the point of 100% good logistics - there's no bonus for unnecessary supplies), whether AGC's ATL October lunge (i.e. Moscow) succeeds is a matter of which tendencies predominate:

• 1. OTL vs. ATL force ratios
• 2. OTL vs. ATL logistics
• 3. Fortification differential.

...and given that the OTL AGC October lunge was such a decisive victory, it would seem that (2) would have to be dramatically, fundamentally different to make the ATL AGC October lunge go fundamentally differently. To make it explicit instead of implicit, I'm analyzing AGC's October lunge as if it doesn't matter whether this occurs towards Vyazma or towards Moscow - but for logistics and fortifications (thus the upthread questions on fortifications).

So I'd argue that meets my burden of production regarding the Moscow lunge logistics, whether I've met my burden of persuasion will be decided by your reply.

I suggest we proceed stepwise in this fashion both as regards Ostheer steps and in analysis of the logistics of those steps.

Next steps would involve post-Moscow logistics, were Ostheer to get that far. As a preview of those proceedings, the per70/TMP discussion would apply: Railheads have advanced farther east during good weather, improving forward rail logistics.

how does that square with History Learner's insistence that the Soviet Union would have nothing with which to defend Moscow after the first kessel?

These discussions are so much easier if you discuss my ATL with me and History Learner's with him. As I've been pretty clear, I believe a one-step lunge at Moscow would have been a poor choice for Ostheer in October (as Halder et. al. recognized) or in September (as I'm not sure they would have recognized).

------------------------

A lot of my tendency to frustration on this board relates to the way I like to discuss things - stepwise in discrete analytical chunks - versus how things sometimes go here - furball with swirling, overlapping, non-specified arguments. After years of doing things in the analytical manner that (good) judges demand, I have lost all aptitude and patience for the furball way. I ask your indulgence.

I have been discussing as if the above explanation of ATL October logistics is obvious - that they'd be roughly as good as OTL or at least not sufficiently worse to turn the October lunge from decisive German victory into defeat. That was, in hindsight, not the best assumption.

[historygeek2021]

I.e. at what stage of ATL proceedings do you see the logistical issues becoming dispositive?

I brought up the issue of logistics in response to your suggestion that the OstHeer would be able to inflict 2 million additional permanent casualties on the Red Army (on top of the Red Army's OTL permanent losses) by conducting encirclements during the winter of 41/42. I doubt that the OstHeer would be able to accomplish any significant encirclements during this period for multiple reasons, one of which would be the poor logistical situation the OstHeer would find itself in during the winter of 41/42, which would only be marginally better than the OTL and perhaps worse due to being even farther from their home base.

It seems you have now dropped the suggestion that the OstHeer would accomplish significant encirclements during the winter of 41/42, and instead attribute the 2 million additional permanent casualties to greater OstHeer successes in the September/October period.

Nevertheless, the logistical situation of the OstHeer in the winter of 41/42 in your ATL is still an interesting question to contemplate, and I think per70 has addressed most of the main considerations.

[TheMarcksPlan]

I brought up the issue of logistics in response to your suggestion that the OstHeer would be able to inflict 2 million additional permanent casualties on the Red Army (on top of the Red Army's OTL permanent losses) by conducting encirclements during the winter of 41/42.

I gave you a list of ATL PoW's additional to OTL:

(LIST)

Adding up all the foregoing gives something like 2mil more PoW - assuming L'grad and Moscow fall.

This didn't include any winter Kessels - mention of that was offhand and is not necessary to reaching 2mil additional PoW. I can't emphasize enough that Ostheer took >5k PoW/day whenever it advanced, regardless of Kessels.

[ok you seem to have the list and judge that I'm walking back the Kessel thing. Not rewriting post though.]

Despite the offhand nature of my winter Kessels remark, it remains a feasible possibility. For example: https://en.wikipedia.org/wiki/Battle_of_Lyuban
While the Lyuban pocket wasn't liquidated until spring, it's not at all clear that this was due to logistics rather than to AGN not possessing the combat power (relative to Soviet) to seal the deal. What proves that logistics was the determining factor?

Likewise, from our favorite author's Retreat from Moscow:

Model had
the equivalent of almost two enemy armies68 operating behind his lines,
which he recognized was as much an opportunity as it was a threat. If he
could successfully cut off these forces, he would effectively turn the tables
on the whole Soviet winter offensive, not only averting the encirclement of
half of Army Group Center, but also inflicting a completely unexpected
counterencirclement on the overextended Soviet pincer.

...Stahel sees the potential for counter-encirclement here... Model ended having success in the ensuing battle but didn't quite have enough at hand to do a full envelopment - that came with Operation Seydlitz in the spring.

Here's AGC's northern sector on Jan 20 (from Stahel):



...obvious opportunities to cut off Soviets there, not just in Model's sector (9th Army). Again it's a matter of relative combat power, not just logistics (though of course logistics impact combat power). AGC had sufficient supplies to fight off the this offensive, including with repeated counterattacks by mechanized forces. Had the Soviets been weaker, as in ATL, it's likely these counterattacks would have pinched off Soviet penetrations rather than merely containing them.

It seems you have now dropped the suggestion that the OstHeer would accomplish significant encirclements during the winter of 41/42

It's not that I've dropped it - just not a necessary condition for 2mil more PoW being feasible. For above reasons, I find it likely that ATL Soviet offensives would have been disastrous, had it sustained the additional Kessels I've discussed (Moscow, L'Grad/Volkhov, Kharkov or similar in AGS sector).

-------------------

I can hear you asking about the logistics of a possible Ostheer winter offensive in January/February/March. It would be, as I called it, a "slow push" so not extending very far from the forward railheads (which are farther east than ATL). Ideally for Ostheer, these would be seesaw battles east of Moscow/L'Grad/Kharkov, with the Germans moving forward ~100km, Soviets countering with their next crop of reserve armies and driving them back a bit towards their railheads, Germans moving forward again after the reserves immolate themselves.

---------------------

Like all fully-mobilized combatants, SU did not have 2mil men standing around picking their noses. Extra soldiers would have come from somewhere, GKO 675 tells from where they came when Stalin wanted 625k men beyond the normal age-class induction process operating in the background (and that, again, would be operating in the background of this ATL anyway).

Are we in agreement on this point?

As ATL implications for '42 come into view, we'd have to look at Soviet force levels and whence comes its regenerative power. That'd involve looking at ATL:OTL territorial holdings and associated economic/demographic resources.

I haven't mentioned territory yet because the GKO 675 dynamic is somewhat of a tradeoff against territorial loss analysis: Because all production in lost territory ceased (obviously), SU typically called up every single military-age male from threatened districts (many avoided that call though and went under German rule). In that case, more GKO 675's may be unnecessary - provided one accounts for production loss under the territorial/demographic analysis.

---------------------------

Doing these ATL's is hard, thus my reluctance to go all out on this one.

IMO the biggest critique of this ATL is simply that the Germans didn't do as I suggest in part because of path dependence from arrogant delusions of quick victory. In this case it's mostly Halder et. al.'s ridiculous concept of the joy ride to Moscow, poof of smoke, no more SU. He/they were too laser-focused on taking Moscow and batting away any non-Moscow action to view the field adequately. Absent help from his generals, Hitler was not a sufficiently subtle thinker to do other than swing a sledgehammer at the nail in Ostheer's flanks.

[historygeek2021]

Despite the offhand nature of my winter Kessels remark, it remains a feasible possibility. For example: https://en.wikipedia.org/wiki/Battle_of_Lyuban
While the Lyuban pocket wasn't liquidated until spring, it's not at all clear that this was due to logistics rather than to AGN not possessing the combat power (relative to Soviet) to seal the deal.

I forgot about that. Good point. The Soviets did overextend themselves in the winter of 41/42, which enabled the OstHeer to conduct some encirclements or near encirclements.

What proves that logistics was the determining factor?

I listed logistics as one of multiple factors that would limit the OstHeer's ability to conduct kessels in the winter of 41/42. I never said it was the determining factor. I was thinking in terms of planned offensive kessels (e.g., Minsk, Kiev, Vyazma), but you correctly point out that the kessels could also result from the Soviets overextending, so I concede that additional kessels are possible in the winter of 41/42.

I can hear you asking about the logistics of a possible Ostheer winter offensive in January/February/March. It would be, as I called it, a "slow push" so not extending very far from the forward railheads (which are farther east than ATL). Ideally for Ostheer, these would be seesaw battles east of Moscow/L'Grad/Kharkov, with the Germans moving forward ~100km, Soviets countering with their next crop of reserve armies and driving them back a bit towards their railheads, Germans moving forward again after the reserves immolate themselves.

My concern is that Hitler and Halder would be overly ambitious and push farther than logistics could support, as in the OTL. However, the timing of your ATL might prevent them from committing such an error through sheer dumb luck - they would likely take Moscow just as the fall mud would make a further advance impossible. On the other hand, once the ground freezes in November, they might lunge for Yaroslavl and Vorenezh and probably even Gorki, putting the OstHeer in a vulnerable position when the Soviet counteroffensive begins on December 5.

Like all fully-mobilized combatants, SU did not have 2mil men standing around picking their noses. Extra soldiers would have come from somewhere, GKO 675 tells from where they came when Stalin wanted 625k men beyond the normal age-class induction process operating in the background (and that, again, would be operating in the background of this ATL anyway).


Are we in agreement on this point?

As ATL implications for '42 come into view, we'd have to look at Soviet force levels and whence comes its regenerative power. That'd involve looking at ATL:OTL territorial holdings and associated economic/demographic resources.

I haven't mentioned territory yet because the GKO 675 dynamic is somewhat of a tradeoff against territorial loss analysis: Because all production in lost territory ceased (obviously), SU typically called up every single military-age male from threatened districts (many avoided that call though and went under German rule). In that case, more GKO 675's may be unnecessary - provided one accounts for production loss under the territorial/demographic analysis.

The GKO 675 figure seems anecdotal. It's just one example of where reserves where called from at one moment during the war. And I'm still not convinced that your ATL would result in significantly greater Red Army permanent losses than the OTL. There would be fewer losses at Kiev, and probably fewer at Vyazma since there would be no Bryansk pocket. And there would be fewer Soviet losses at Leningrad (and if Leningrad isn't encircled, its population and industry can make a greater contribution than OTL). The big potential for greater Soviet losses is if there is a second big encirclement at Moscow in early October. There we also have to factor in Hitler/Halder's proclivity for spreading the schwerpunkt out as thin as possible, so it's possible they don't even take Moscow in October if Hitler and Halder are diverting the panzer groups to Kalinin and Tula.

But, even assuming a successful and significant encirclement at Moscow, I don't think there's enough evidence to show the Red Army's manpower reserves would be weakened to the extent necessary to win Germany the war. If there are 8 million personnel in the Soviet armed forces instead of 10 million in Spring 1942, or 4 million frontline soldiers instead of 5 million, Germany is still in the same strategic dilemma as the OTL. And we know what Hitler would do - he would drive for the Caucasus ... and now you're going to tell me to slow down

[TheMarcksPlan]

I don't think there's enough evidence to show the Red Army's manpower reserves would be weakened to the extent necessary to win Germany the war.

Neither of us has done enough work to say anything smart on the issue. Given COVID vaccination and the return of warm weather in my parts, I don't anticipate doing so.

[History Learner]

I don't think there's enough evidence to show the Red Army's manpower reserves would be weakened to the extent necessary to win Germany the war.

Neither of us has done enough work to say anything smart on the issue. Given COVID vaccination and the return of warm weather in my parts, I don't anticipate doing so.

If I may:

IV. Remaining unused resources:​
a) reserved for employment in the civil economy - 2 781 000​
b) in labor columns - 1 321 000​
c) recruits born in 1925 - 700 000​
d) non-conscripted men fully fit for service in the Central Asian Military District - 600 000​
e) non-conscripted men with limited fitness or in the age above 45 (without Far-East and Transcaucasus) - 500 000 (of them 277 000 in the Central Asia)​
f) non-conscripted men in the Far East, Trasnbaikal and Transcaucasus Fronts 505 000 (including 200 000 with limited fitness and 200 000 in age above 45).​
g) officers of reserve, not conscripted yet - 156 000​
h) expected convalescents from hospitals in 3 nearest months - 350 000​
i) in the penitentiary system - 1 156 000 men in age from 17 to 45.​

​
About the only source left is the 700,000 men in the 1925 age class and the 600,000 in CAMD, but the latter are of questionable utility for political reasons.

[TheMarcksPlan]

If I may:

...and in an ATL in which SU has lost more territory (Moscow region, northeast, perhaps everything up to the Don), it also would have lost many from the '25 class and from older age cohorts listed in Schadenko's document (Art's source). As Schadenko states, SU lost >4mil military-age males to German occupation up 1.9.42 excluding Bessarabia and the Baltics.

There's absolutely no way to get 2mil more men in '42 without taking them from reserved occupations. Whether the administrative route is a GKO675-style direct levy on the reserved categories or via mobilizing the entire male population of lost areas doesn't really matter: If the latter then the mobilized don't join their evacuated factories in the Urals; if the former they come directly from the factories. Either way it's a serious output decline in '42.

GKO 2100 was a smaller repeat of GKO 675 (100k instead of 600k) and was less heavily-weighted towards the basic industry workforce than was 675 - by then SU was in crisis regarding the metals-fuel-transport sector and could not take as many from it. Nonetheless, a quarter of the draftees came from metals-fuel-transport and the armaments commissariats gave up 2% of their workforce.

Now multiply that levee by 20 if SU needs to replace 2mil additional Barbarossa losses: a production collapse of 40% in armaments industry (certainly more, given the productivity level of reserved male workers).

...so obviously SU couldn't have replaced 2mil more Barbarossa losses for the '42 battles. There'd have been some compromise between poorer arms and smaller army. It's feasible that W.Allied desperation could have made up some of the deficit with even greater LL deliveries but hard to see how much. Given the shipping crisis conditions of '42, a doubling of LL probably would have required shutting down the Guadalcanal or Alamein offensives, especially considering the post-PQ17 routes were long.

---------------------------------------------

If History Learner and/or HistoryGeek2021 - or anyone else - want to make a rigorous argument regarding the impact on '42 military outcomes, I'm all ears.

The essential point is that it's profoundly unserious to look at these manpower and economy constraints and pretend either that the situation is unchanged or that German final victory in '42 would be a cakewalk.

[History Learner]

If History Learner and/or HistoryGeek2021 - or anyone else - want to make a rigorous argument regarding the impact on '42 military outcomes, I'm all ears.

The essential point is that it's profoundly unserious to look at these manpower and economy constraints and pretend either that the situation is unchanged or that German final victory in '42 would be a cakewalk.

Personally my thoughts have always been that a 1941 where Moscow and Leningrad have fallen is one where the USSR starts to come undone in the Winter of 1941-1942, and then 1942 sees the Red Army gradually collapse over the course of the year; probably with the Germans achieving the A-A Line or something close to it while 1943 is them moving up to the Urals. Probably no formal treaty, but what's left east of the aforementioned mountains just isn't really a serious military threat. I'd imagine the U.S. goes Pacific First in 1942 and then you get peace either after the Republicans win the midterms that November or FDR finally bites the bullet in 1943 when it's clear the Russians aren't coming back.

[KDF33]

...so obviously SU couldn't have replaced 2mil more Barbarossa losses for the '42 battles.

and then 1942 sees the Red Army gradually collapse over the course of the year

IMO, there's very strong evidence that the USSR, in its rump state of late 1942, was unable to cover its historical losses, to say nothing of those postulated in this ATL. First, let's look at Soviet mobilization data for 1943-5:

11,098,929 call-ups, of which 6,226,600 (56.1%) came from the RSFSR and 4,872,329 (43.9%) came from the other republics. Note that over this period ~1,073,600 men were demobilized and sent to work in industry and a further ~817,600 were sent to staff paramilitary formations, most prominently the NKVD. Obviously many men were also drafted from industry and the NKVD, so overall those ~1,891,200 men are effectively swaps. Thus, the net manpower addition in 1943-5 is 9,207,729.

This report suggests at most 1 million men were still mobilizable in the Caucasus and Central Asia by late 1942, including those of limited fitness and above age 45. If we're conservative, at least 4 million men must have come from liberated federated republics. To them must be added men conscripted from liberated RSFSR territory in the winter and summer of 1943 - how many is unclear, but given respective population pools I'd say another million is conservative.

This would leave ~4 million 'net conscripts' from the rump USSR of late 1942, which compares to 3,203,400 men conscripted from the new age classes of 1925-7.

IMO, by late 1942 the USSR had to start recapturing territory to make good its historical losses, to say nothing of hypothetical losses on the scale of 1941 or spring/summer 1942. This is ultimately unsurprising: the Germans mobilized just 3,565,000 men between 1.6.42 - 30.9.44 out of a population of ~80 million as they conscripted millions of foreign workers. That the Soviet rump state, with a population in the ~120 million range and no access to foreign labor, could have done much better should be patently absurd to everyone.

The Soviets' success along the Don in late 1942 and early 1943 has obscured how precarious their position would have been, from a manpower perspective, had that offensive failed. In the event, success literally fed upon success, all the way to the pre-war border and, shortly afterward, Berlin.