I think you make a good point with respect to Winter 1944/45 and Spring 1945. For pretty much all of the other turns the 'beat history' concept works just about right, I think.
The challenge as you converge to a win or lose finally in Summer 1945 was the design to make the two VPs for both Axis and Allies converge at a pretty even rate through the last turns of the game. Also, it was important to keep a delta between one side winning and losing wide enough and converging through the last turns of the game as well. For example, there is a 10 VP margin to keep the game going on turn 20 (16VPs vs. 26 VPs). This converges to 8 VP gap, 6 VPs, and 3 VPs in the last few turns of the game. - forcing a victor by turn 24 with a 0 VP gap.
There are also a couple things that are not modeled perfectly in the end-game. Like, in the game, the Germans would never pull out of Athens - there are no partisans. More likely, in the game, the Germans may fort down in Athens and try to hang onto the VP there at all costs.
Also - I think you need to 'force' the Axis to try to hang on to more VPs than historically late in the game - or else they will just give up places and fall back into 3-4 fortress spaces and try to hang on to the end. In reality, the Germans fought desperately for places like Budapest, keeping an army in Norway, and launching a huge counter-offensive to try and retake Antwerp. So, I think the inflated VP level for the Axis late in the game tries to model in that level of desperation a little bit and also try to stop the Axis from 'gaming' the system and just falling back to rapidly and not fighting for some of these locations.
That being said, I am up for making adjustments to the game - one of the great things about it is we can take ideas from people around the World and implement them. Can you propose a VP table that you think works better and post it here?
thanks,
mark