The problem is not a technical nor architectural problem but a user/usability issue.

Look at the workflow how people create a new community. They are registered on one instance, probably fixed to that bubble and probably don’t interact with other instances at all (subscribing to other communities is a pain. Other problem.) They might (if at all) search for a similar community on their instance. If they don’t find one, they’ll create a new one. Searching every single community is not implemented in this flow. You need to call up feddits search to do so.

My suggestion: Either do a name (fuzzy) check when creating a community, listing the ones already existent on other instances. Or at least implement the search feature from feddit.

I'm confused. Are you one of the lemmy devs? Have you talked to them about prioritizing this? If not, who's going to make it work? It's all well and good to propose things, but unless someone is going to be able to code it up and integrate it with what's already there, it's kinda like pissing in the wind.

I'm not trying to piss in your cornflakes, but you're proposing weeks to months of work that may be moot without stability and scalability being solidified first.

Or is this a call for developers to jump into this proposed project, and I'm misunderstanding?

That's ridiculous that only agreements they made are with apps that look and feel at best like apps straight from Android 4.0 and iOS 7 respectively.