I believe that gender differences exist, because (among other things) I believe in Darwin's sexual selection theory.
I wonder why Morocco and Tunisia are the best counterfactuals that authors can trot out. For one thing, they are rather distinctively different in culture from much of the educated world. For another, they only achieve approximately 5% and 10% the GDP per capita of the U.S.
I have a theory that math abilities can converge in lower socioeconomic subcultures, and diverge in higher socioeconomic cohorts. The findings of this study would seem to support that theory.
I think there's some more wood to chop here. But I am ready to eliminate math ability as a gender difference.