I can live with badly rendered counters if the basic consim is good. My problem with UV and WitP, and also most other consims, is that the research is shit. It's insufficient to know the names of ships or how many guns were in Unit X if you don't have the least clue about the results those things tended to produce.
WitP is married to the "EXP" model of combat resolution, and that model is not only NOT well researched (the EXP values are arbitrary Axis Fanboy nonsense), it's also not a good model to begin with. The lesson from WW2 is that SUFFICIENTLY trained persons with SUPERIOR tech consistently and far more often "win" against massively trained/experienced persons with inferior tech. The other lesson is of course logistics, logistics, logistics, to which WitP and most consims pay mere lip-service, except where they get it flat out wrong.
Shit strategy/grand-strategy WW2 games that I have owned, played, and given, thrown, or traded away:
Advanced Third Reich (AH)
World in Flames (ADG)
"Advanced" WW2: ETO (Decision)
A World at War (GMT)
War in Russia (GG/SSI)
Of those AWAW was the worst. It was idiotically conceived, idiotically written, idiotically detailed, chose an idiotic time scale, and idiotically tried to blend the look and feel of tactical combat into 3-month game turns. The economic and research models are idiotic. The political models and assumptions, when they have any, are idiotic. The combat mechanics are idiotic. Most of the stats (combat value, move, range) are idiotic. And the details are most often completely wrong when it comes down to specific named ships.
I have found one good grand strategy consim for the ETO: Totaler Krieg (Decision)
I have found one relatively good grand strategy consim for the PTO: WW2:PTO (SPI version... not "Advanced" PTO by Decision Games).
That's it. Based on the ignorance that is so obvious among many game designers, it's a waste of time hoping for someone to write a game that gets even the basic relative power projection relationships right.
Against my better judgement . . . could you be more specific?
I agree that, many games (WiTP included, and to a somewhat lesser extent WitPAE), in many genres do not seem to have even tried to validate their modeling and this is a shame because it would seem that it shouldn't be that hard.
Set up tests with AI vs. AI. Run tests, repeatedly. See if results fit with expectations. If they don't adjust mechanics.
If you start that process at the beginning of design, and start with the least inclusive elements and validate how things work at the lowest levels first, then work your way up . . . well then, you've pretty much followed the same principles as how mechanical or electronic machines are "engineered" and that would seem to be about the best that could be done.
I blame it on the developmental roots of video gaming: pong, tetris, space invaders, etc.
From the outset, computer games were not meant to be "simulations" they were meant to be beguiling, and it seems that few developers have really shaken that ethos. With good reason too, because a game that truly achieves a high threshold of naturalism without ALSO achieving a sufficient threshold of beguiling is liable to be a commercial bomb. Just like any business, designers and publishers have tended to take the "less risky path."