Personally, I want to be able to make this assumption, to have this trust. Because the A/B balance seems to be tipping.
But it's so easy to say "I feel Less Safe when you say X". It's only human to notice that some words and phrases take on power. Only human to want to use that power to protect yourself.
So instead of owning your own feelings, you own that of everyone in the group. And online, that's everyone. Some people online will claim injury even when there is none, or when they imagine they're speaking for others who are staying silent.
The result is a discourse so sanded-down and smooth that it achieves only the distinction of offending no one.
Isn't the issue that it allows bad actors to pretend to take offence and we don't have a mechanism of validating whether the offence is real or malicious?