There were also glaring issues with the statistics, as outlined in Andrew Gelman's blog[1].

[1] https://statmodeling.stat.columbia.edu/2020/05/25/this-contr... (one of the many posts there)