I don't see how this personal take generalizes to a trend specific to data science (as much as the term looks weird to me). In other words, I don't see how creating useless Python notebooks is different from over-engineering or creating useless features.

To me the proper context of the story is "cheap money have been flooding the market for _decades_".