Recent items by Richard Mac Manus (on single-topic blogs) and David Weinberger (on the future of blogs) raise in my mind the question of the nature and boundaries of a Web Log (or Wiki Log) (Micro Content). A given person writes lots of chunks of stuff: in his own blog, in other blogs' comments, in email, in review sites, etc. Other people do the same. There are many ways to "package" (publish) a collection of those bits (single-author/single-topic, multi-authors/single-topic, permanent vs temp group, permanent vs temp topic (or Cate Gory), Blog Thread, etc.). How does the meaning/relevance of a given bit change as viewed as parts of different collections? How do you (why should you) keep track of the various contexts in which your various bits appear? How do you even keep track of the various bits that you write?
Update:
This somehow also remind me of the recent re-thread about "What is a Web Site?". In some logical sense, the concept of a Site isn't useful, because in a way the entire World Wide Web is one big web. If we're talking about documents that are logically related, you could argue that the separate postings by 2 different bloggers about the same event or 3rd-party article are more "related" (in a topical sense) than 2 sequential postings by the same blogger. Or, conversely, that a blogbit I write and a related comment I post on someone else's blog are more related than 2 sequential items in my blog. I think the folks playing with this stuff need to distinguish between a Site as a logical collection (or collection of collections) and a Site as the set of all documents served by a Web Server (note that even this is fuzzy since you could have a single piece of hardware running 1 or more installs of a Web Server software package hosting a number of Sites).
(Oct22) Richard Mac Manus notes the distinction between blogging as Publishing vs as Communicating.
(Dec8) Paolo Valdemarin on decentralizing comments and group blogs.