I've got to hand it to Dave Winer and how he's able and willing to tweak some of our basic assumptions of what a blog looks like and how it works. Fundamental things like, "a blog is a reverse-chronological set of postings." I mean, there goes my short-hand elevator pitch definition. But just being willing to take a totally fresh look at something that has become established is a wonderful thing. Kudos!
This experimentation encourages us to broaden our thinking on what a blog is, how we use them, and what role they have in the overall world of the web and open writing. (Hmm... WWW = World Wide Writing?)
It also makes me re-look and think again about the relative roles of the web rendering of a blog and what I see in my aggregator. If crucial information is available only on the site (like categories, frequently updated topics, etc.) then suddenly the value of my aggregator diminishes.
Yes, perhaps the special features might have their own RSS feeds. But that doesn't solve the problem. It's not really practical to follow more than one feed for a single blog. There are ideas on how to handle this, for example a taxonomy of rss-feed types, or a set of related feeds, with extensions to RSS to represent them.
So bottom line: what are the relative roles of the web rendering and the rss rendering(s) of a blog?