All good observations Flame... And I agree; grains of salt required
It all started with news surfing and I ran across this:
http://blog.aarp.org/2015/02/27/why-women-should-check-a-mans-finger-size/?intcmp=AE-BL-IL-BLOG
Which led me to this abstract:
http://www.sciencedirect.com/science/article/pii/S0191886914006400
And this one (which suggests there is likely an answer to your question):
http://www.earlyhumandevelopment.com/article/S0378-3782%2804%2900012-X/abstract
After that, I poked around a few other places, and it appears the correlation regarding hormone exposure in vitro is well documented... Some of the other attempts to extrapolate social habits, sexual orientation, gender identity, etc likely show so much scatter that truly valid correlation is unlikely.
I decided to post the Wikipedia entry because it provided a broad overview (at the expense of accuracy)... And it comes with a handy chart
I agree, before I buy this one fully I want to see some real data... Population, sample sizes, scatter, etc.
-j