ORIGINAL: Anthropoid
In short, why did ACW battles happen where they did? Since there are not that many major ACW battles (60? 90?) it might not be that hard to simply categorize each one by one or two categories of "WHY IT HAPPENED WHERE IT DID." Lets say you come up with three categories (a) Opportunistic; (b) Strategically-Important location; (c) Choke-point.
What would Gettysburg rate as then? Shoes can hardly be ruled out under any of the abc's you listed [:D]
For every battle that fits into category (a): why bother with creating a map. Create maps for the other two only. Some provinces might have no (b) or (c) type maps on them. Some might have two (b)s (Strategically Valuable Locations) that are linked by one (c) (Choke Point). If you win the scouting, you can choose to force the battle at the b1 site. If you win it, then you are in a position next turn(or maybe in the same turn, depending on number of troops, and their mobility) to try to take the (c) battlefield site (the choke point that links B1 and B2). Win the battle for the choke point site, and you are then in a position to take the B2 site . . .
Will not this force you too keep on attacking no matter? I usually play the South, and there are times when I am satisfied just throwing those Yanks back into the Potomac, and having no desire to press the counterattack. Reason tends to be one out of two, a) I need to rest, resupply, and bring my numbers back up. b) I need my forces to stay on the defensive as the enemy would otherwise invade the area I am currently in.
The other interesting thing which this approach could facilitate is to build off of the existing AI tendencies which some of us have complained about a tad bit (e.g., your comment about repetitive Union AI attacks on Fredricksburg). Right now, the AI has a tendency to just attack anywhere it can, which I think a few of us have argued is not very realistic. If certain provinces were more strategically important as "doors" into additional areas, or as highly valuable strategic caches of some sort, then the AI might not behave quite so ahistorically.
Well is repetetive AI attacks on Fredricksburg that hard to understand? I would not move the Army of Potomac south unless I could force Bobby Lee out of Fredricksburg. Sure, I might bypass him and head for Richmond, but then he is free to head north and take Washington, and we are back to square one.
I think what might be an issue is that the AI focuses on destroying the enemy forces in the field (Montgomery comes to mind) while most players I guess goes for territory (Rommel comes to mind). So I do not think you would need to make some areas more important, but maybe change some parameter for the AI?