My opinion ...
Day of Infamy: Starting in Sept. 7, 1941, Japan's probability to DOW the U.S. should be 25%, increasing by 25 percentage points each month to 100% on Dec. 7. Any DOW automatically triggers the Pearl Harbor maluses and the Day of Infamy boost in recruitment.
The Japan DOW should also trigger scripted invasions of Philippines, Guam, and Wake. The benefit of the US holding Guam and Wake is way-way OP.
If Japan wins in China, it should have a ridiculously strong navy, with every mighta, coulda, shoulda in history built.
The opening decision to give the bonuses to the Allies should include an option to deny those bonuses to the U.S.
The proposed increases in MP should at some point include a tradeoff in productivity and/or demand for consumer goods (which is really the pacify-the-masses index).
If possible, certain strategic choices of the U.S. might bring bad press, i.e. dissent. Convoy raiding, for instance, had been successfully portrayed as craven in the Atlantic while simultaneously pursued by the U.S. in the Pacific. Perhaps not just being raided, but being the raider should include a small unity hit.
The point is, the U.S. needs to be much harder pressed.
Some other points on the wishlist: A way to automate strategic bombing. The bombers get wasted so quickly, this mission requires too much micromanagement. But if I put an HQ in London, attach bombers to it and automate it, my planes always get relocated to some far-off, home location. Is it too gamey to give the U.S. some little piece of territory or European cores or whatever it takes to trick the AI into blitzing Germany?
Some bonus to underground cells in France to simulate the French Resistance. Currently, IP invested in the underground is pretty well wasted because a Fascist government can find and destroy them so easily.
This is with all humble obiesance to DD.