[tokenadult]

Thanks for sharing that link. The underlying paper is (of course) better than the popular article kindly submitted to open the interesting discussion here. In particular, the paper makes clear its own methodological limitations.