1. I do follow snake_case for naming my files and folders
2. All files should go into corresponding folders by their categories. So all the books will go into the books folder and so on
3. When I have things in the folder that I don't use now I archive it inside that folder so that it is not cluttered (If I have some pdfs that I don't need now it gets archived)
4. And temporary folder for things for things that will get deleted later (at the end of the day)
5. No deeply nested folder structure beyond 3 levels since it will become harder to search for things (except for node_modules and other software stuff)

I recently found about this sub-reditt reddit.com/r/datacurator/

I'm switching to PARA method. Barely started to change the structure, but I'm really looking forward to it, cause my current system is horrible. i've been using PARA for my personal knowledge system.