Recursive RSS
We usually associate RSS with blogs and news sites, which is fine, but it's also a narrow way to look at things. It might be more helpful to think of RSS as a description of a "channel of information", where each piece of information ("item") has a title, URL, and a short description. There's a little more to it, but that's pretty much it. You can see how it lends itself to news-like text-based items, but that's not a built-in constraint of the format.
One guy who's almost always thinking about things in the right way is Micheal Sippey, and in his blog he mentions the ability in Feed Demon to represent some kind of filtered aggregation as its own discrete RSS feed. You can do a similar kind of thing with Feedster searches, where the results of an RSS search can themselves be expressed as an RSS feed.
Once you've agreed upon the relatively simple structure for information exchange, it opens up the whole food chain to innovation: producers (sources of RSS), consumers (RSS clients), and, perhaps most interesting, the middleware (aggregators and filters like Feedster, BlogStreet, and many others).
When we developed Spyonit, we had a number of ways to be notified when your "spies" fired: you could get an email, an SMS, an instant message, or it could be reported on your "My Spy Reports" web page. If I were to build that today, I would most certainly add "as an RSS feed" as a notification option (not too mention the ability to monitor RSS feeds for alerting purposes ... but that's obvious [smile]), so that your spies could report to you through your favorite RSS client or aggregator.
Stay tuned to this channel ... I think the RSS world is going to get more and more interesting.