People aren't uniformly distributed among the different types either. I accept your point though; the distribution of programmers across Meyer's Briggs types isn't the same as the distribution of all people across Meyer's Briggs types.

Correlation is not causation either though (not that you said it is).