Anyway, preference is a preference, you cannot explain it away. (Analogically, if a male prefers females with big breasts, you can't change his preference by explaining that bigger breasts are not necessary to feed children.)
What I meant to ask was what kind of information about someone is hard to detect in a few hours of face-to-face interaction but would still affect someone else's willingness to have a (usually very-short-term) sexual relationship with them, regardless of whatever evolutionary reasons caused such a preference to exist. (So, in your example, the equivalent “men like women with big breasts” would be a valid answer, but the equivalent of “men like women who could produce lots of milk for their children” wouldn't.) And I didn't mean that as a rhetorical question.
(FWIW, and I think I've read this argument before, it would make evolutionary sense for women to have different preferences for one-night stands than for marriage, because if you had to choose between a healthy man and a wealthy one you'd rather your child was raised by the latter but (unbeknownst to him) had half the genes of the former.)
I think humans have general preference for "reals value" as opposed to faking their reward signals (a.k.a. "wireheading"). Of course sometimes we fake the reward signals, because it is pleasant and we are programmed to seek pleasure; but if we did it without restraints, our survival value would go down. So when someone enjoys "fake values" too much, they will get negative social feedback, because by putting themselves in danger they also decrease their value as an ally.
So a part of mechanism that warns women against "fake...
If it's worth saying, but not worth its own post, even in Discussion, it goes here.