If your culture makes it hard for people to enter, then demanding that people prove they belong to it before they can comment on their difficulties is an immunization strategy.
Demanding strong evidence for the difficult to measure impact of social habits of the in-group on outsiders is a good way to eliminate debate.
What he is saying is reasonable for dealing with discrimination within the scene (and particularly point 5 is very important). It is useless when talking about the place of the scene within wider society. Why are girls less likely to develop interests in computers? I doubt that this has anything to do with actual sexist judgements within the scene. But I do believe that it has a lot to do with the overall social norms of the scene being aligned with male gender roles. Ideally we just get rid of gender roles, but we are far from that. In the meantime we need practical solutions.
Finally and most importantly: The idea that the open source community is so meritocratic as to be immune to biases that have repeatedly been shown to exist, for example in academic mathematics, is ignorant at best. Wilfully so at worst. It shows an anti-scientific mindset on this particular question that belies the "show me the evidence" rhetoric.