I'm in the process of coming up with terminology for various theories, similar to lots of other work on LessWrong and The EA Forum.
Naming things is a bit of a unilateralist action. While community members don't have to accept a specific naming proposal, they are likely to do so if they like the concept. I can't think of many cases where Eliezer or someone named a concept, and the community decided that that name was poor, and renamed it.
However, I can't find much theory on how to figure out great names for things, or even what to consider when doing so. I would have expected there to be comprehensive discussion around Information Architecture, UX Design, or the Library Sciences on this topic, but couldn't identify much outside of card sorting and a few lists of rough heuristics.
This was also an issue for me when I did more software engineering, and I was then also frustrated by the lack of discussion I could find. The best there was work on Software Patterns, which I used primarily for naming conventions.
Some related links I could find:
https://en.wikipedia.org/wiki/Naming_convention
https://www2.staffingindustry.com/eng/Editorial/Archived-Blog-Posts/Adam-Pode-s-Blog/Probably-the-best-file-naming-convention-ever
https://www.invisionapp.com/inside-design/naming-conventions/
https://ux.stackexchange.com/questions/48578/naming-features-of-an-app-or-site
https://www.martyneumeier.com/strong-vs-weak-names
Kind of famous programming quote (Martin Fowler credits Phil Karlton for it, but it's likely much older):
Really, naming is idea compression in hard mode. You need to find very short strings that encompass your current and future desire for this category. In a semi-adversarial environment where other people will misinterpret your meaning by taking it too literally or not literally enough.
Their clustering of ideas is slightly different than yours, so the label will "naturally" not align across any two people, and may really hit different clusters across some. It's going to take tens of thousands of words to debate which is the "true" meaning, and the original namer isn't really in control of what ideas win.
I very much recommend Habryka's and Petter's idea: don't start with naming. First think about idea organization and transmission. Some things you should probably NOT name, in order to avoid compression artifacts.