1. They actively encourage removal of material that is accurate

While I mostly agree with your argument here especially [[Spoo]], if one edits the articles Jimmy watches the chance of you getting a mailing list post or a usenet group post source to stick is very low, even if it isn't documented anywhere else.

For example (just one of several :)), I ran into this problem trying to explain a key part of the [[Merkey]] history as there is sort of a "reliable source blackout" on some periods that were clearly "notable" to the subject.

So, perhaps that is the "de-facto policy" there.

I think that our problem may be that, because we place such a great demand on our sources, people don't bother to source articles at all. Perhaps we need to demand less in order to achieve more

I disagree and IMHO anyone who is a fact-tagger for a while on more mundane articles will see the issue quickly.

Often, when I would {{fact}}-tag something, someone would remove the tag and instead go into a five-page essay on why a certain point was "correct" or not. Usually it was, but sometimes it clearly was dubious; and usually needed attribution anyway. It is difficult to attribute something when you don't know its source :).

The good part is that usually they just go "OH!" when you let them know that all they needed to do was source the thing; which ends up with a solid source almost every time :).

Maybe this has something to do with David's theory about the encyclopedia being written mostly by anons/new users (my own theory is that it is sort of 40/60, but anyway). I'm assuming at some point, if this is true, then there will be some kind of software measure to make sure something added is sourced.