"And while this is somewhat true, it gets the causation backwards. It was obvious to me from a very early age that I was better at sight reading as a singer, so I spent more time doing it."

This is self-refuting really. You admit that you spend a lot more time on one skill and then report that you are better at that skill and are hopeless at the one you tinkered with.

K Anders Ericsson, who is one of the sources of the research on this speculates somewhere (I can't recall of the top of my head where), that there are very likely extremely small marginal differences in initial aptitude, but that these alone are enough to garner praise and tiny edges in performance that are self-rewarding to the practitioner and that these advantages make them more likely to continue with the work required to get better. That's not natural talent, if it's true, that's just a minuscule advantage that may help foster interest and practice and lead to better skill.