>>/4890/
If I remember correctly, there used to be a significant correlation between autism and homosexuality, which diminished as homosexuality became more accepted. The likely reason was that autistic people are less likely to hide it, not that the actual rates are different from neurotypicals.
I don't think I'd put my confidence as high as 50%, but it's completely consistent with his hypothesis for it to be visible in weird places without truly being more prevalent in those places.
There's an implicit assumption that the environment only affects visibility, not the underlying condition.