I think this situation would be better for me if the bookmark manager would download the page, if possible extract the main content part sans the navigation and teaser boxes and built a searchable full text index from that.
Optionally a snapshot of the page could be useful. For that I am misusing Zotero and the Zotero FF plugin. Apart from managing papers I have to read I also keep track of informative web pages.
With sing Buku, emacs integration instantly came to my mind. Could be useful with the W3 browser mode or other modes that display URLs.
Main reason I keep so many tabs is to send them to people when it comes up in conversation
That'd be why I'd consider buku anyways
Also in the back of my head is auto-pushing bookmarks to wayback and replacing the bookmark link with the wayback copy ala gwern.net; presumably doing that with sqlite would be simpler than some kind of chrome plugin
I find that even with a couple of folders of bookmarks in my browser, reading something later just doesn't happen in practice. And for reference sites that I might need later, the browser bookmark system is more than enough. Is it just me?