Back to seriousness for a minute... I wonder if a lot of this is hurt by the innate imprecision of language. It's all very well stating that "women are more likely than men to overthink" or what have you, as the study may well be true, but it is clearly and obviously glossing over the fact that the overlap is sizeable in almost every case.
I am not sure that this point is ever really emphasised properly. The memo in question acknowledged it, vaguely, before going on really to ignore that in pursuit of whatever else it was trying to say.
I'm happy to accept that there are differences in men and women overall, as an average, but it has to be taken alongside the massive variations between men and women themselves. And that was kind of what prompted the question I had to start with. How can biological differences alone account for such wide disparities as a 4:1 ratio among computing engineers? How can it be that, even if men are more predisposed to become computing engineers on average, that we still pick more of the "above average" men than the "above average" women?
There are simply more factors at play in these gender gaps, in most fields, than can be accounted for purely by any biological differences. Even if those differences are "only" being exaggerated by society then it's worth acknowledging that. Ironically, one of the reasons is that indeed men and women can bring different perspectives, life experiences, etc. to the job, and therefore employers could proportionately lose out on this variety if they are missing some diversity in the workforce.