ORIGINAL: Sonny
What is really needed for historical accuracy would be either to incorporate the actual problems and difficulties that made the Union victory take so long or gameplay restrictions/balances.
The first would be almost impossible without a super computer and would be far outside the scope of this game - or would confine the player to historical paths which give the player so few options as to not make playing enjoyable. The second seems to upset the grognards.
Supercomputer? Since when has a supercomputer been needed to play a game?
The art of game design is to represent the essentials of a complex situation in a much simpler way. If you tried to represent all the factors of the original situation to ten decimal places of accuracy, sure you'd need a supercomputer but that's not game design. The fault of almost all modern computer games is that they include
too much detail -- because they have the technology to do so. All this detail just gives players too much work to do, the games take too long to play -- and the essential strategical factors in the situation are obscured, because the game hides them in a welter of unnecessary details.
For instance, in a game of the whole American Civil War I see no need to represent all the different weapon types explicitly. There were various different weapons used at the time, yes; but I submit that you could arm all the troops with the same weapons and it would make no real difference to the strategical situation. The quality of weapons could just be factored into the quality of the troops in general.
There are two main tests of a game like this. As I said at the start of this thread, one test is whether you can pursue historical strategies and get roughly historical results. The other test is whether the game stops you from pursuing strategies that would have been impossible in reality. (When I say 'strategies', I mean grand strategies affecting the whole course of the war, not little exploits such as cavalry raids, which made no real difference overall.)
If the game passes both tests both with and without a certain game feature, best to discard that feature (or least make it optional), because it's complicating the game without contributing anything essential to it.
I don't criticize a brigade-level game for not representing regiments or companies. That's a valid design decision and surely a game can pass the historical tests without representing small units explicitly. Probably a divisional-level game could pass both tests tolerably well without representing brigades either.
I've love to see an ACW game that could be played to completion in a few hours while passing the historical tests. I believe it could be done. But no-one seems inclined to produce such games. Sigh.