Traditionally, feeds (RSS, ATOM and RDF) have been used to distribute a website's content. Subscribe to a site's feed and you get the "headlines", a stream of items usually composed by title, publication date, abstract and main body.
What most web developers know is that this simple "schema" fits well enough not just "news" but other kind of data too. Take for example Flickr "photo streams", where instead of a "body" you get the URL of a photo, del.icio.us bookmarks where each item is a user's bookmark, or podcasts where each item "encloses" an audio or video "attachment". One of the nice things about RSS, RDF and ATOM is that they are flexible enough to support uses like the above and (by design) are not limited to "news" items. Nowadays, almost everything is published as a feed -add wish lists, alerts, personalized search results and much more to the above mentioned.
Users tend to think that a feed derives from a web page content, when usually they are both representation of the same information that reside in some sort of database. The main difference is that usually web pages are focused on presentation, when feeds are focused in structure. As mentioned above, a feed has an inherent structure that makes it ideal for other programs or services to "consume": parse, understand and extract "what matters" per case.
FeedBurner already gives some of these services. You can easily "mix" your blog news, with your Flickr photo stream and your del.icio.us bookmarks. You can even mix them in different ways, say gather all your weekly del.icio.us bookmarks in one single post in the new feed generated. Or you could create your own "video channel" of videos indexed by Yahoo! and tagged with a certain keyword, like "funny", or "football". Or automatically post your del.icio.us bookmarks as a single post in your blog on a daily basis using yadd. Or combine a GeoRSS blog feed with geo-tagged Flickr images. Or upcoming.org events with rsswether.com...
I'm quite sure (and Burning Questions' article point to this direction too) that we will see more of this "rip-mix-burn feeds" trend in the close future. We should expect to see new tools that allow us to extract, parse and combine information from feeds into new feeds, as well as presenting those new feeds in new ways, generating unexpected results, services and added value.
-- UPDATE. Check out this article too: The Second Coming of Content and RSS Feeds