If I'm allowed to cast my vote, I am nether in favour nor against a different fatigue behaviour.
I am much more in favour that the developers and scenario makers get it exactly right, which means to model it as historically and phisically (or should I say psycho-biologically) accurate as possible.
This may mean to review the fatigue mechanisms itself, but also to carefully check how all the relevant factors (morale, cohesion, experience, training, fitness, ...) interact together.
I can, though, confirm one thing Franklin said: with an ordinary (no "fastest") move order to a HQ, its subordinates seem to often be force marched. Again, this may or may not be right, and maybe the force marching periods are just short intermediate tasks used by the HQ to put the units in the formation in order not to have the whole battlegroup stay on prolonged wait.
On the other hand, to say something on GBS's remarks, if you pick a scenario with very fit and trained troops (for example play as Allies in the Tutorial St. Vith scenario), you'll notice how much more responsive and snappy they are throughout, so I assume fitness and to some extent training are two remarkably important factors in the overall fatigue modelling.
Again, I really don't know, so I also cannot tell if an ad hoc rule of thumb or a quick&diry tweak of adding 10% of fitness to all units would do.
But I trully would love to be able to rely not only on the fact that the "AI will do a reasonable job" (as a motto of the game says), but also that, while doing that, the AI operates in a reasonable world.
Keep on the good work.
< Message edited by johndoesecond -- 11/25/2010 1:32:21 AM >