To me, the choice of Git repo should be one of the more portable aspects of a project. I know there are hideous lock-in mechanisms like issue lists and wiki discussions, but hopefully someone will standardise them one day into Git objects.
1. People already have accounts, so it lowers the barrier to reporting issues and contributing.
2. GitHub provides some level of discovery that's hard to replicate otherwise.
Both of these are a function of GitHub being popular. It doesn't have to be a particularly great tool if it's what people are already familiar with. I wish it were otherwise, but I don't see a real alternative that can compensate for these advantages in the short term.
I love your idea of having issues/discussions/etc managed directly in Git—that seems like a more flexible and adaptable system. Even if somebody developed a standard like that, though, it seems impossible to get any of the big repo hosting sites to adopt it since it directly competes with their own value proposition :(.
I think this was Oxide and Friends show, but I'm unsure.
Makes sense on its face, popular projects = potential money. Whether that's a good signal or not is a different discussion. I even think on the episode they mention how misguided this was, people using something might not mean there is potential for revenue if the people are using it solely because it is free and licensed appropriately.
People are just looking for patterns anywhere to justify whatever they want. I've seen npmtrends used to justify adopting certain libraries. Really makes you wonder.
The Shopify acquisition of Remix (https://news.ycombinator.com/item?id=33405997) established the viability of this investment strategy.
Stars are a vanity metric but, they can help with monitoring some early growth.
Everyone on this thread should list their projects, then we all star each others stuff all the way up!
If enough people act like it has value, eventually it kinda does.
Github is a package registry for NPM, Docker, Ruby Gems, Maven, Gradle, and NuGet. Quite a lot of things would break if it disappeared.
Mailing-list likers have not come to grips with this in the last twenty years and I don't expect them to start now, but the dismissive attitude might be preventing you from learning something. "Principal Skinner, it is probably not the children who are wrong."