Shaun
In the main, just three simple folders.
- Work data (stuff worked on at present)
- lookup/reference data (data for lookup sth. at a time)
- Archive data (archive)
What's in there is actually conceptually self-explanatory. As few sub-levels as possible. No file may be located on a folder level. Strict convention on how to handle and name (new) content. Result: Finding without searching.
Thoughts:
The desire for less friction honours you. My credo is simplification, which results in speed and time savings. I think a problem only arises if you want to combine some kind of mind management (such as e.g. the F✱ N approach in Notes) with a file system).
Regarding PARA, I found the concept idea and what Thiago wrote mildly logic. But it wasn't for me because it clashed with an existing, effective document management system. Such ‘wisdom’ on data structuring is only ever beneficial if there is no performing system in place.