I'd like to see data on academic performance in specialties. Most well-paying first world countries employ specialists (eg, doctors, programmers, lawyers, accountants, etc). Recall Simpson's paradox in statistics, where a general distribution doesn't have to apply to the individual sub-sets.

FTA:

> Women who go to university are more likely than their male peers to graduate, and typically get better grades. But men and women tend to study different subjects, with many women choosing courses in education, health, arts and the humanities, whereas men take up computing, engineering and the exact sciences. In mathematics women are drawing level; in the life sciences, social sciences, business and law they have moved ahead.

This paragraph seems to support the idea of my above musing.

> academic skills have no bearing in getting high paying jobs

I highly doubt that there is no correlation, but I doubt that education is the sole factor. Girls performance over boys is a slight, but significant, edge, but since success likely draws from a handful of other factors, a few of them having a counter slight, but significant edge, might be enough to swing the other direction.