I see what Tyler is getting at - crudely put, a level playing field between otherwise isolated demographics will encourage an efficient exchange of goods and services, which means that beautiful women of all social backgrounds will congregate more efficiently where rich and poor meet.
Now, assuming that beauty is at least partly a function of genetics, this still has implications for the attractiveness of specific ethnicities. (Yes, I realize this is a slippery and subjective slope so I will avoid concrete examples.) If you're a woman from a culture that has been urbanized for longer, chances are better that these disparate demographics have been able to arise and then mix in the "marketplaces" that Tyler is describing. Consequently, women from long-urbanized gene pools will have had more time, pressure and opportunity to produce more beautiful combinations.
Now think of the opposite case: if you look like Angelina Jolie but you happen to have been born in a hunter-gatherer village of 100, then sure, your one in a million looks will get you the "best" guy (best hunter, OR chief's son, OR best-looking) out of that hundred. But the chance that you'll land the best on your continent (the male one-in-a-million, i.e. Brad Pitt) are low. So, among hunter-gatherers (or more recently urbanized people) you would not expect the same selection for beauty as a huge population urban center with admixture.
Note that this assumes a model of women valued by physical appearance, men valued by social status/wealth/power, a situation which the more enlightened among us will find distasteful but also recognize as having obtained for the vast majority of history (and probably all of prehistory). Consequently the same theory could also be used to argue for males having more of whatever it is that equates with wealth and power, if a) there is non-hereditary mobility within the urbanized society (i.e. the high priest's son can lose his spot if someone memorizes more potions than him), if b) the selected-for traits are constant over time, and c) if those characteristics are genetic. Those are thinner assumptions (especially the first two), but maybe explains why attractiveness in males and females don't seem to go hand in hand.
To test this, we could start naming cultures that we would expect to have more attractive women, but the only way to test this idea is against some universal index of attractiveness (i.e. women from group X on average are more attractive to Zulu, Inuit, and Clevelanders). We do not have such data, although it wouldn't be that hard to obtain. Consequently, at this point we're just arguing based on personal biases and current taste.
Like Tyler, I too spend a lot of time thinking about location theory.
A study of the geographic forms in cartography
7 hours ago