I'd partially push back on that one in that I think that experience is very much environment dependent, because what I'm seeing from LLMs in a professional webdev environment isn't all that special. Even good LLM code tends to get pointless comments inserted that just repeats information that very readably is laid out in e.g. the variable name it describes. 5-line functions whose required information for understanding is pretty self-contained in the 5 lines you need to write them often enough also get doc comments attached that arent just pointless but actively harmful because they're not going to be kept in sync. More can be said about other things that LLMs do some of the time that arguably isn't good practice.

And I've got to review similar levels of code for years before the LLM train hit the timeline. I wouldn't have said the people are obsessive, that's just... decent code that you write to be readable.

Now in a science environment where people have barely even heard of coding standards and readability in myths and legends, I'd agree with you. But that has more to do with the average scientist turned coder having absolutely atrocious coding style that was a travesty even before LLMs came into play.