The methodology of the blog post you point to repeatedly is laughable. The main objection is that the cross-sectional correlation the author claims is completely incompetent to show causation.
I'm also not sure what causation you believe Sanandaji is attempting to show. Could you explain?