They're not impossible, they don't mean give up and go away, but from my experience and gut, I agree with Anonymous Funder that these challenges are real and they will take a lot of thoughtful, hard work -- and figuring out how to sustain that work long-term in a volunteer community dynamic.
If people want to run their own Cohost(s), which I would personally love, or future sites like it, then I highly encourage you to get together and do work like @invis has started doing on the legal and organizational aspects, which are always underrated, as well as the technical architecture problems, which are difficult even without adding federation into the mix. Open sourcing and licensing are important decisions, but largely easier and separate from the big strategy of how to run a technical organization and platform with random users who will depend on you indefinitely (and malicious users, too). I'm glad to see people proactively engaging with unsexy but crucial stuff like legal sponsorship.
As an analog that people may or may not like: People on my Mastodon feed spend some time talking about specific technical issues with the ActivityPub protocol and the server software, but a lot more time talking about how to manage the workload, moderation, interpersonal, and legal challenges of running your own Mastodon instance; how to govern the Fediverse at large; and because of all of the above, how to avoid big shutdown or defederation events that mean confused users have to scramble to move their stuff. Hell, there is very cool funded research that you should read whose only ambition is starting to get a handle on the best current Fediverse governance practices and future needs. These are Big Human Problems whether it's an open project or not, even if you just wanted to self-host a very popular phpBB.
I would be delighted if enough folks have the wherewithal and skills for it, but I have endless grace and understanding if they don't, or if the time's not right (it's not right for me, either), or if you put your efforts into improving other existing communities who also definitely need the help. If the whole basket of skills doesn't come together, then it's not Cohost 2 (at least not for long); it's just a sparkling raspberry-colored website. So I'm glad people are taking pains to anticipate these problems, or to take a point someone made and really engage with it and come back with a better solution. If it doesn't pan out, you did your besht!
