Safety on social media is inherently biased. Somebody or some group has to make that call. That's why moderation can't be done with respect for free speech at the network level. Doing it only at the personal level means people will be exposed to crap until they see it and block it themselves. Instance and personal control therefore is a great compromise, but it can't make the whole network as "safe" as some demand, especially when they try to compare the #fediverse to platforms like old #Xitter.
It's not necessarily true that users will be exposed until they block. Alternatively, they could only see feeds that they allow.
So some demand one thing, others demand the opposite, and empowering users to make those decisions as to their own experiences is a great way to go.
