It's funny, but I quite often create "expert" articles on subjects I'd bet that few wikipedians know about, or have even heard of (e.g.:< http://en.wikipedia.org/wiki/Cauchy%27s_equation%3E). But I've yet to have one listed on AFD because they are ignorant of the subject, and I'm in no danger of being driven off. Perhaps there's more to the problem than just the "proud ignorance" of AFDers.
The problem is definitely much less severe in technical fields. But all of these specialized articles serve a common problem: failure to establish context.
When I wrote [[conditional quantum entropy]], I didn't include enough information to tell a general audience how important it was. It might be taught in every undergraduate QM course (no). It might be a specialized topic used by specialized researchers (yes). But people see the article, and say, "this is science, I can't tell how important it is, I don't want to be the one who nominates this". I'm confident that it can get filled out later.
But in the current atmosphere you can't do that with, say, webcomics. Any article on a webcomic *should* explain to a general audience what the importance of the article is. Things are important because of their larger context, and a webcomic article needs to explain that it has 1000 readers, that it was the first keenspace comic to use this-and-that artistic technique, that it was a major influence on Penny Arcade, that it was an early exploration of this-or-that theme in webcomic history. All of those things are much more important than a list of characters, but they're *hard*, and maybe they'll come over time. But these days we're not waiting for them.
In conclusion, I have no idea what position I just took.
-- Creidieki