This tells me that, at least in the case of sexual preference (and possibly, all other preferences, such as in learning how things work/STEM), it seems to be significantly (but not completely; but also not insignificantly) influenced by nonsocial/biological/"root" factors.
This is of course, unfortunately, the most troublesome answer, because all it does is lead to even more questions.
But this only provides evidence of a genetic component to career choice, not sexual differences. To prove that, you'd need to isolate the genes responsible and show they are on the X or Y chromosomes. Actually, genes are a lot more complicated than that, and there would likely be a large number of interactions between a lot of different genes. So instead you'd need to use some sort of statistical model.
Also, I'd note that this doesn't rule out societal factors. Heritability scores are a measure of the interaction between genes and the environment. If we lived in a world where it was against the law for a woman to be a computer programmer, you'd see a very strong genetic component for career choice as well!
I don't think that's necessary. All you'd have to do IMHO is
1) take male identical-twin-raised-separately data about career choice
2) take female identical-twin-raised-separately data about career choice
3) do a multivariate regression analysis of both twin-to-twin and also gender-to-gender; you should be able to derive the statistical relationships between career and sex, and both of these to genetics, and that would demonstrate genetic influence.
Since the correlation between genes and career seems to be positive http://www.aei.org/publication/the-genetics-of-job-choice/, and the correlation of gender and career seems to be positive http://career.iresearchnet.com/career-development/gender-and... (although is claimed to be 100% environmental, according to the blank-slate hypothesis, which I consider a radical position), we can safely assume that some combination of all 3 is also positive, although the actual value requires math.
Looking at modern-day gender ratios in careers and just hypothesizing a biological cause is specious reasoning. You need so much more evidence to make that claim.