Thanks to Jeff and Tony for their comments, my reply comment got long enough that I figured I'd just make it a post.
Campaigns are definitely a good way to go.
Having thought more about this, scenarios for le petit geurre usually have built in victory conditions so its mostly pitched battles or the typical, if historically rare, encounter wargame that the question applies to because the object is just to beat the enemy unless its part of a campaign or a scenario with set victory conditions (such as control of a road junction or the establishment of a bridgehead).
Running through a mental list of various pitched battles it seems like typically, at the end of the day, either:
a) 1 side has collapsed. In wargame terms some sort of division, corps or army morale failure or a preset limit in figures or units lost. Lacking these the game becomes b) by default.
b) both sides are still capable of fighting at the end and the decision to stand or retreat is a calculated one by the general based partly on the condition and relative position of the 2 armies and partly on the strategic situation. In wargame terms a draw when time is up unless part of a campaign or having an arbitrary victory condition either in terms of losses or holding key roads etc.
So in game terms, I am ok with leaving win/lose conditions to be part of the scenario design (even if it is just destroy the enemy's ability to fight) but I think there should also be the possibility of "army" collapse built into the rules.
Right now my horse & musket rules have an optional rule forcing an army to give up if 1/2 of units are broken or destroyed but they used to have rules that allowed for a form of army collapse by means of morale tests and penalties as brigades got hammered. Seems to me now that I think about it that the 1/2 rule discourages reserves and I can't remember why it replaced the other. Needs some pondering.