I just wanted to say that I still think the article doesn't really show surprising to people who took an undergrad stats/data science course, and the apparent 'conundrums' are well-understood.