Mastodon has proven that these moderation issues can be solved. The question is if the Lemmy devs (and community) can politically agree that Mastodon-like moderation tools are a good idea to prioritize. That's where I'm worried.
I guess its not a big deal because there's bigger fish to fry right now, so to speak. (Plenty of simpler bugs that everyone agrees upon). But eventually, the fundamental user-tracking / karma-system / vs defederation / design of anti-trolling tools becomes political and ideologically based. What happens to the community then?
Hard for me to say. I hope that the overall community accepts the problem and agrees its a problem worth fixing (even at a slight tracking / anti-anonymity features getting pushed to Lemmy-core). If not, maybe the problem can be solved with bots (aka: IRC) that helps track Karma-like scores and whatnot across servers and helps auto-moderate content that enters a particular server or even community/sublemmy.
I think I can "imagine" solutions to these problems. But what I can't foresee is what the community will deem an acceptable solution or not.
-------------------
Mastodon is a very good model though and has led the way by showing what kind of tools are needed to curate a community. I see that Lemmy kinda-sorta wants go to about things a different way though.
Lemmy isn't a "fork" of Reddit, as much as it was just a Fediverse thing that existed and happened to become popular in-and-around the Reddit blackout.
The Fediverse-model is good. Mastodon is good, but a poor-fit for people looking for things like Reddit. I think what Lemmy is, and what the Lemmy-users want Lemmy to become, remain an open-question.
Some of these problems also apply to kbin.social. I expect different "Reddit-like Fediverse replacements" to spring up, implement competing solutions, fork, die out, etc. etc. over time. This is the nature of all things open source.
The only way to figure things out is to write code and try it out. Its not like anyone really knows what they're doing here, its a lot of guess-and-check and experimentation. And that's fine.
I think we've more or less recognized that communities built around volunteer moderators have needs that Reddit has failed at (and likely will fail again in the future). Building alternative solutions is an open question in general.
Lemmy for now, remains the forerunner. But I can imagine Kbin.social taking over if they make the right decisions (or if Lemmy developers "pulls a Reddit" and forces another migration).
I think the Lemmy.world instance has very good administrators who have made good decisions thus far.
But as for "who to follow", that's a bit more ambiguous. Its not necessarily important to follow people you agree with... its sometimes more important to follow people you disagree with.
Lemmy.ml IIRC is the Lemmy developer's instance and is therefore important, even if their posting styles do not match what I like to see. I also consider Beehaw.org important, they rub me as snowflakey but I think I respect what they're trying to do (even if I don't necessarily agree with it). sh.itjust.works errs on the other side politically.
The other fun thing about federation is that when you ban an idiot coming in from another server, you never have to see them again and they can post their last word out there in the ghost version of the thread that you never have to think about.
Implementation details do matter however. I could see something like Kbin.social winning out as PHP-developers could be more popular / available than Rust developers, for example.
Hard for me to look into the future. Weird decisions ripple out over time. People are definitely "onto something here". So I'm trying to keep an open mind to the different possibilities.