From: London, Surrey, United Kingdom
Your gut feeling is correct. I completely agree that the speed of entrenchment and the effect that fortified units enjoy is just too much in TOAW3. I believe that Daniel Mcbride struggled on this very point when trying to convert DNO to 3.4, alas, a project that is now abandoned. We know these things through Lzard (Kurt, where are you?). Even in 3.2.29, the Soviets are compelled to build a Maginot Line further back. When the Germans reach that line, the game degenerates into trench warfare. I've played the Soviets once and I've been soundly beaten (by Fungwu), because of my urge to defend too far forward, so I wouldn't call my self an expert
Well, the Soviets had basically their entire army destroyed in the first couple of months of the campaign. If they are able to preserve it and give it time to rest, prepare and entrench, then absolutely it should be able to stop the comparatively diminutive Wehrmacht.
However such an action was as impossible for the Red Army as all their soldiers suddenly developing X-ray vision and super strength. It would have required a level of awareness and foresight which was totally lacking, and in any case would probably have led to the utter political collapse of the Soviet Union as sixty million citizens are abandoned without a fight.
Simulating a major military debacle is always a massive design challenge (I wrote a paper on this back at university). One needs to understand why the losing party made the decisions it did, and attempt to place the player as much as possible in the same position. Ultimately, though, most people playing a scenario like this will be fully aware of the relative strengths and weaknesses of the two forces. Perhaps the best solution would be to start the scenario six weeks into the campaign when the fate of the Western Soviet Union is already sealed.
I've never made any serious attempt to simulate this campaign so I'm not the one to answer these questions. However, they do need to be answered; otherwise you're going to end up with something which might resemble history, but certainly doesn't simulate it.