As a vim user, I too only use plain-text. My general writtings are organized in folders, within a single git repo. Project specific writtings are within one file within that project's repo. I always call it notes.txt, and then when I'm within a project, to open notes it's always the same operation (`:e notes.txt` or equivalent). I also have a notes.txt in my general writtings repo. It's by far the biggest file in the repo, where I keep information that can be useful in the future. In most files, especially the bigger ones, I keep to the structure of date stamping a day's entries. For that I use a vim abbreviation. That's 1 line of code and that's the only custom code that my "system" needs.
I have found it immensely advantageous to use monospace for writting and note keeping, because I can structure the writting with indentation. So my writtings are trees, I branch them similarly to how you punctuate text in English. That's very efficient for technical writting and notes, but also for ponderings and essays. It let's you drop many non-content words, because the indentation provides for sentence-glue with visual punctuation and walks you through the "thought tree". With prose it's not easy to see the structure from a glance, you only see it's biggest features. With lists, you are constrained to the point where the medium is not expressive enough for anything beyond a grocery list. With loose-form trees, you get the benefit of visual structure, text-efficiency and expressiveness. Code editors, like vim, are good for this, because they know how to wrap indented text.
Example of such notes. They're not in English, but you can see the structure. It's a week of entries on a project: https://hastebin.com/raw/gurexecubi