That’s a great question. I just keep them in my vault. Ahrens’ model, IIRC is to have lit notes only be lit notes, write fleeing notes as you read that cite the lit note, and then turn those fleeting notes into atomic notes and delete the fleeting notes. Lit notes, fleeing notes, atomic notes are all kept separately rather than being integrated into a single storage medium. Am I remembering that right?
Personally, I just leave my lit notes in my vault, and write my fleeting notes directly into the lit notes. After I finish reading something, I look over my fleeting notes and decide which ones should be split off into their own atomic notes, which ones should remain in the lit notes as notes on this text, and which ones should simply be deleted.
The second of those three options is where I depart the most from normal ZK methodology. I’m a text guy: I think by reading, and I spend as much time studying particular texts, eras, or authors as I do thinking about ideas in the abstract. Taking in other people’s thoughts and thinking my own thoughts are inseparable processes for me. So I need to have the ability, not just to connect ideas to each other, but to connect texts to ideas and texts to texts. Having permanent lit notes, with not just a citation but my own running commentary on the text, linked to atomic and MOC notes and integrated into my vault, is the solution that works for me. YMMV.
This also means that I don’t really tag my lit notes in any special way (other than the title being the BibTex citation key)—I have no reason to ever want to distinguish my lit notes from my “own” notes (e.g. in a graph), because my lit notes are my own notes.