To be clear, I'm going to show all the battalions, regiments and batteries in my big battle games. In three rather short sittings I've completed almost 230 French battalions already, and I only need another 100 for all the battalions at Leipzig. I have about 50 or so regiments of cavalry, and 50 batteries- I need about 24 or 25 batteries. Then on to the Allies. But I digress.
This project is really about the visual appeal; the game needs to be fast and sort of simple to run well (since there will be upwards of a thousand units on the maps, and a lot of players pushing them around with croupier's sticks). So, although the units might be moved around individually, the combat resolution is going to have to be abstract.
I have an idea for quickly resolving large level division combats in a game of the scale I'm thinking about. The first idea is that the turn is one hour. Each division represents a "power pool," from which each turn (an hour) the player can deploy a certain number of "power points" for an attack. Each battalion model represents a certain number of power points. The division commander can choose what percentage of his unit to attack with and risk each turn (alternatively, he may be constrained by a dice roll of the maximum amount of units he may risk, as determined by conditions and psychological factors of the commander). Attacking is done by scooting the chosen battalions forward towards the enemy position to be attacked. The combat is resolved, and either the position is taken, the attackers route back to their parent division, or the combat bogs down into a stalemate (that can be easily broken by whichever side is quickest at hurrying in reserves). This is all designed to be rather loose on the table, and t require no exact measuring, which won't be feasible in a game this big if anything is going to be accomplished.