After posting "Shopping Carts for Content" on a IA discussion board, I received a lot of responses that were interesting. The following is a digest of what people (mostly IA's) had to say.
Many project have attempted similar approaches to content access. The closest to what I have described is International Herald Tribune. And yes, it is called "clippings" but still does not offer the ability to email all the "clippings".Financial reason for *not* providing this service: content providers want to increase traffic and for users to pay for services. For example NY Times, WSJ and Salon.com now charge for premium and archived content.
Technical reason for *not* providing this service: providing static links to content that is dynamically generated is a hassle to maintain.
One person suggested going further and not limiting the user to yet another pre-defined tool by tracking information elements to the paragraph level to provide greater control. A good content managment system would allow the user to choose which content to capture and how to use it.
What I had specifically had in mind when I was thinking about this service where sites that provide large amounts of text documents that users would use primarily for research purposes. Large governments sites and sites like the World Bank fall in this category.
Even so, I don't see why paid service sites like the NYTimes couldn't provide this service within their paid areas.
I guess one of the issues may be whether to create this service at a site level or as an independent app not tied to any given site. The International Herald Tribune is an example of the former, and iLOR is an example of the latter. I would advocate the former for the time being, since it allows for the content provider to format the information in the way that is most appropriate for their content (on the assumption that they know their content the best).