I’m the guy you quoted above (Greg Hill). I’d like to preface my remarks by pointing out that this article was written from the point of view of individual bloggers setting up navigation systems, rather than an analysis of a site from the entire community’s perspective.
1)
“Categories are fewer in number than tags”. I still stand by this. Just because a few tags become popular for a particular site doesn’t mean the unpopular tags somehow don’t count. Blogs that employ the “pure” tagging approach (ie not tags-as-categories) typically have dozens or hundreds of tags. You might be able to find the odd pathological case – but I doubt it.

2) “Categories persist for a long time, but tags don’t”. I could have been clearer. I meant the “active life” ie still being associated with content, not just used for retrieval. This is in contrast to the “one-shot” tags.

3)Generic vs specific. If you’re using “news” as a tag, I’d suggest that’s a category! Which pretty much gets to the nub of it. The flavour of my post is to define a category as a particular (restricted) kind of tag. Why?

Well, in the context of organising blog posts, some blogging platforms support “true” convential, traditional categories (like WordPress). Others (like Blogger) don’t, and they have to be bolted on with hackery. The major focus on Freshblog is doing this with “web2.0″ tech (tags, feeds, web services, AJAX etc). So, for people doing it this way, all they have is tags. The question is whether they wish to “replicate” categories with tags or use tags “natively”. In practice (as I pointed out and you seem to agree), many people use a hybrid approach, where frequency indicates the “degree of categoriness”.

Cheers,

-Greg.