You don't like pdfs?  

 I’ve been using Calibre to convert epubs and pdfs without images to .txt, and reading them in less.
you can search for any string with / and in less, the v key opens $EDITOR, and you can add a —bkm to search for next time you’re reading, or add “margin notes” in [[ ]] brackets. 

 I have Calibre on my laptop.

Less can do margin notes??? I should make more use of margin notes. How does it save them?

I use Pandoc for a lot of conversions.

And no matter what I use, converted PDFs still have glitches - dropped paragraph initial letters, random spaces, tables turned to mush... 

 I mean YOU add margin notes. You come up with your own system.
For me, I’ll add a (1) in the text of a paragraph, then beneath the paragraph in double brackets I’ll write a note. And when it’s in .txt, all the Unix tools like awk, sed, cat, etc, work. 

 Fair enough! Yes I do that, but thought there might be something integrated and fancy! 

 I get that. Nothing wrong with well built tools.
I just personally have an affinity for old school, manual minimalism. 

 csv isn’t bad if you pipe it through quotecsv and awk -F,
I was screwing around with a single 20mb file with noting but csv text in it last week 🤓