I'd push back on your last point; there's a zero-sum feedback process here that is invisible. You're seeing the negative feedback from a vocal set of minority power users; the positive feedback from people isn't going to be known or seen in anything other than usage metrics.

You say "users": how many, what percentage, what cohort..you get the idea.

Some people don't care, but the ones who do care and don't like the change, tend to really not like it.

Personally, I think the importance of not upsetting the established userbase far outweighs the probability of maybe growing that userbase a little.

My concern is that they'll count any interaction with this new forced editor as a "win" and pretend it's an A/B test when in reality it's a lark.

I somehow doubt a lot of people are going to be happy about a rich text editor that doesn’t work.

A bit like the teams editor, which was forced upon me and is absolutely horrendous.

If we lack percentages, how do you know it's a minority that dislikes this?

How do you tell the difference between unpopular features and ones that are mostly liked? In both cases, only "a vocal set of minority power users" would say anything.

With the state of statistics literacy in this industry, "usage metrics" often mean whatever the person citing them want, with no malice or deception intended. Accidental, inadvertent p-hacking is shockingly common.