I treat links like atomic notes. I add as much detail as I feel like to each link, sometimes I go back and add tags and notes. Then I have an exceptionally poor process that attempts to go back to each link, get the archivebox archive and uses python to attempt to grab the article text (I tried using newspaper3k at first, but it’s unmaintained, so moved to readability). Then sticks the resulting link text into the note.
Honestly It’s a mess, and I really haven’t figured out how to do link things together very well, but, for now, it’s my little disaster of a solution.