Your first argument discounted psychological studies primarily on the basis that there are two many dimensions to account for in relation to a measurement of behavior.

You then went on to suggest a link between gender identification and the results of a study on a single variable. It’s probably the case there are many environmental and physical dimensions at play with gender identity. While I can’t say at this point that your link is wrong, it may be best to hold back on the last assumptions; especially a biologically based one, given the scope of the impact.