I did a very short 2 week test with the 'quick fix' (due to the way I'm writing my current Soviet AAR I have voluminous back records of the original game), so I chose a relatively contained sub-section of my original game (put the rest on AI) and as far as possible reran the original sequence of attacks (it was a major Soviet offensive and the Germans on that sector were clearly on the defense). Looking at the final casualties with the new values, any difference (for both sides) is well within the normal margins of variation (I think Soviet losses wre around 15,000 as opposed to 16,000 and German around 22,000 as opposed to around 21,500) for land combat. What happened that was different was losses due to air bombing dipped from about 6,000 Germans to around 2,000.
So my strong suspicion is that the in-game values are perhaps the reason why sustained airstrikes in SF can be so damaging (not just to your vital support brigades). I offer no view as to which should be correct, but my feeling is that airpower should do bad things to organisation and movement speed (incl advance/retreat in province) rather than be a major cause of direct casualties.
In effect, this returns us to where the debate has always been on these two numbers. Its clearly not what was intended but then the entire game (esp the combat) system has evolved around those values as currently set. Change them and it probably will over a longer game, with a greater variety of combat interactions, and by changing the air-land dynamic, seriously change how the game plays out, but from my little test, its pretty easy to see how no one went looking for the problem that was wrecking the combat system (as it simply wasn;t).