You can already block people on Twitter. The problem arises with shared blocklists. How is a shared blocklist identified? Well, the odds are that it will probably need to have a name. You now have the following problem: the name can be used to promote hate if the name of that blocklist is visible to other users, or to falsely associate a given user with other nefarious groups, as Google will probably crawl the lists and the results will show up in searches. The whole thing ends up being exactly an added moderation mess, just like what you started with, but with a few more layers of indirection and different ways it can be abused. Plus you still have the original problem of moderating messages that needs to be solved.

Solutions like this look great initially if everyone uses them properly, but everything falls apart when people inevitably start actively abusing the new feature. The design needs to handle assholes-at-scale from the outset.