That's not the argument. I think the argument is that when people are told that they are expected to do worse than their peers, they do, but when they are not told that, the gap disappears. Therefore arguments from statistics about innate ability to perform being associated with other characteristics (being a woman, for example) when widespread, become circular.

This is the argument.

>[...]belief in gift and natural born ability is toxic to all, and by eliminating it, we won’t alienate so many from technology.

This is a consequence of the argument.

As to the identification of some actual physical ability in math or programming which is defined at birth, it requires a positive proof that's not circular. A study on the innate relative unemployability of American blacks would find a lot of statistics to back it up, but would be a circular waste of everyone's time and a consequence of that (based on experimental evidence) would be a higher rate of unemployment in American blacks. Note that the fact that there's a pragmatic consequence of pursuing a field of study in a particular way has no reflection on the truth of the thesis, yet taints the effort of determining that truth.

As far as I'm aware, the only arguments for innate mathematical ability are dripping with survivorship bias.