or maybe some other terminology would be better? lots of people get confused when you ask them to choose an instance, sometimes I think even the word "proxy", "host", or "hub" is simpler
the specific terms aren't my point, just a discussion to see if we can come up with a better name
High Availability stans are angry about your question.
Basically... both options are ambiguous. Would be best if they used a more unique name. Like a burrow (many lemmings build burrows).
I kinda don't hate the idea of coming up with our own unique name like Burrow, feels like something Apple would do lol. But then it wouldn't be a common term across the fediverse, and also I feel like it would require explanation and then you'd have to choose how to explain it anyways, and you'd still have to use a word like server, instance, portal, provider, etc.
Both sever and instance have multiple meanings when it comes to deploying Lemmy though.
An instance is running Lemmy publicly, but also just running the APU creates an instance of that API. To scale, you'd probably run multiple instances of the web API.
Same applies to server, but worse. You could also call the web API a web server. You could also call a VM a server. You could also call the physical machine a VM is on a server.
When it comes to naming stuff, it's best to find something unambiguous if it's a core defining thing you want to tell people about. Private corps do this to build a "brand", which is still a valuable thing for open projects so that they can gain adoption.