This goes beyond what Google Reader did and would provide a feature really seldom available on the web (I know only of Pinboard, which is a paid service).

If a feed article didn't supply actual content (i.e. picture links, sites that only give out snippets, or shared links from other RSS feeds), I'd like to be able to archive the linked website so that I can see the content in 20 years from now, as well.

Should probably be optional and the archive should be independent of the overlying implementation of the shared items (see this other milestone).

Needs some discussion, obviously.