The fact the the authors of the paper rush for innate explanations (hormone levels in the womb...) that they have no expertise to comment on instead of much more obvious behaviors (maybe self-labelled gay and straight people upload different types of pics or take care of themselves differently, on average), is a clear violation of Occam's razor.

The "average gay face" is similar to the "average straight face" with glasses on, a better angle, and a couple of pounds less.