I was writing about the article linked here. But since you mention the memo, I'll say this.

The current state of science suggests associations between gender, brain development, and career preferences. But women who are applying to Google for coding jobs are clearly stating their own individual career preferences. People apply for jobs they want.

Using population studies to try to contradict clearly expressed individual preferences is nonsensical; that's not how statistics works. It's like the old joke about the guy who carries a bomb onto a plane for safety, because "what are the chances that there are TWO bombs on one plane?"

And obviously associative data about preferences doesn't tell us anything about individual capability or qualifications.

Finally, if you actually ask women in STEM fields why they left or are considering leaving, career preferences are not the only answer. One also hears about disrespect, harassment, abuse--the glass ceiling. Again, this is real data that can't be lightly set aside just because it doesn't fit a narrative of biological determinism.