The problem is in the unintended consequences. Once we started accommodating requests to change potentially-problematic words at my last company, those words went from "possibly problematic to someone" to "officially confirmed as offensive by company policy".
From that point forward, many people assumed that anyone accidentally using those words had ill intent. This peaked when an interviewer chastised a candidate for referring to the "master branch" during an interview. We also had someone try to cause problems because our Linux systems had "man pages", which they believed was proof that Linux distributions were sexist. It's one thing to search/replace your documentation for a specific word, but just wait until you have a team of people brainstorming complex plans to remove "man pages" from every Linux PC and server in the company.
Ironically, once we stopped making company-endorsed efforts to navigate problematic vocabulary, the number of people offended or insulted started to decline. The policies had the opposite of the intended effect.
It's one thing to make a personal effort to use a less contentious vocabulary. I have no problem with that, nor should anyone else. However, the problems come when using industry-standard vocabulary is assumed to have ill intent. No one wants to operate in an environment where they can be declared to be racist, sexist, or otherwise prejudiced for using a basic technology term without ill intent.