Two important questions must be answered first.
First, we must ask if there is true correlation between the autism spectrum placement and coding skills. If there is, then it may be reasonable to use autism spectrum location as a screening variable when hiring coders (note, correlation is not causation, so be careful here).
Second, is it harder for women to grow into or be born into that pocket of coding creativity? If it is, then we should expect the ratio of successful programmers to be imbalanced, just as there is clear imbalance in the NBA based on height.
It is the worst sort of unicorn-buying for our institutions to try to "correct" such an imbalance without understanding these aspects of the problem.