I think you are right to suspect the methods and the results. If you look at the paper's github, the python notebook was clearly written by a chatbot (the comme

I think you are right to suspect the methods and the results. If you look at the paper's github, the python notebook was clearly written by a chatbot (the comments are all in the second person). So what you have here is a monograph, unreviewed, unpublished, based on GPT-level understanding of a survey that might not even apply to this subject.