Victorious units will usually have suffered losses which degrade their subsequent combat ability. These losses, although low for a computer game, are probably higher than is strictly realistic. In part, therefore, they act as a substitute for fatigue recording.
I am not afraid to agree that compromises have been made in the name of playability. That is what game design is all about.
In playing the game, it seems as if fatigue is modeled, in the way units behave. Units that have been in combat are less efficient, and seem to move more slowly. There is a morale status that degrades the effectiveness of units, and units are much more likely to break in the later stages of the game than in the earlier ones. It also seems that units can move more hexes in the earlier game than the later game. It is very difficult to get units that have returned from a long pursuit back into the battle, as they seem to have less action points.
While it appears that fatigue may not be specifically modeled, the game plays as if it was, based on my experience.
I agree with is sentiment. What your seeing for game mechanics is:
*units that are fragmented lose the ability to maneuver well, (loss of AP's)
*units with high casualties (under AI control) are very reluctant to close the distance, basically they don't want to autobreak
*some units, especially lights, when fired upon when moving sometimes lose AP or the ability to reach the target tile (basically interdiction fire)
*The # of men remaining in a unit directly effects firepower but also influences melee combat(but not impact)
*units might begin a battle with kiel status, but lose it once degraded via casualties
When its all put together as a whole, yep, the feel is units certainly are tired and more brittle and less affective bear end game, which makes having a reserve ideal.