We're all off to Sebastian Fiedler's place to talk about custom-built RSS feeds. Hope to meet you there, there's an interesting discussion forming.
I had an idea that if RSS is such an obvious format to use for reusable learning object (RLO) syndication then why not use RSS autodiscovery? Put simply, if a web page that contains RLOs has an RSS autodiscovery header item then all of the objects on that page become discoverable by automated systems such as RSS news readers, auto-subscribe bookmarklets and more importantly, RLO aggregators.
So I made a demonstration page. Pages such as this backed by a content management system would make the insertion of the RSS link automatic thereby making resource discovery and the reuse of learning objects fairly painless.