I don't know much about this field (and I did not purchase the full article), but it seems like quite a leap to assume this is a source of the gender-gap. First of all, you're talking about 80% of one standard deviation difference in reading-to-math scores between males and females - that doesn't seem like enough to study it as the cause of anything.
And how does being better at reading steer people away from STEM? There is just as much reading going on in STEM as out, and arguably more intensive reading in.
It seems like the score difference is probably related to the gender-gap somehow, but I doubt it is a cause. I would be interested in reading the article if it was freely available.