I have very little to add to Becker's excellent discussion. One puzzle remains is why women have better college grades than men. One possibility is that colleges discriminate against men in admissions. For if colleges admitted blindly on the basis of academic prowess, they would keep admitting women until male and female grades were equal at the margin. The average grades of women might still be higher than those of men. But this would be surprising, unless most of the students in the applicant pool were women.
Discrimination against women in admission to college would not be irrational if male alumni are expected to be on average more generous donors, either because of higher average earnings or because, as Becker notes, men are likely to dominate the upper tail of the income distribution; alumni in the upper tail are likely to be disproportionately generous donors.
Another possibility, unrelated to current sex discrimination, but perhaps to historical discrimination against women, is "legacy" admissions. If alumni children are favored by college admissions officers (largely for financial reasons--admitting alumni children increases expected donations by alumni), and the alumni parents are disproportionately male because men used to go to college in higher numbers than women, this could explain why males are being admitted who are expected to be poorer students than women who could have been admitted in their place. However, given that alumni are likely to have an equal number of male and female children, this explanation would work only if alumni prefer their sons to be admitted to the same school.
Still another possible explanation for the higher average grades of female than male students is that men get as much out of college as women do even when male grades are lower, because there is more to college than academic performance and the "more" may be more valuable on average to men than to women. Male sports and other male social activities in college may build teamwork, and networks, that create more valuable human capital for men than these activities would do for women, perhaps because men will have greater participation in the labor market, where teamwork and connections are vital assets. On this view (proposed by Asher Meir in correspondence), male students substitute nonacademic for academic college activities, resulting in lower average grades that are, however, offset by the social human capital that they acquire from engaging in the nonacademic activities.
Whether the wage gap between men and women will continue to narrow because the ratio of male to female college students will continue to fall seems to me speculative. The ratio may not fall at all if colleges see advantages in the current ratio, though this would leave unexplained why it has fallen as far as it has already. If the ratio does not continue to fall, I do not see what would drive female wages up relative to male wages. Rising prosperity may actually induce many women to substitute household for market work, because diminishing marginal utility of money income, combined with higher income tax rates at higher incomes, would tend to make untaxed household income more attractive.