The Cunctator wrote:
Both mentalities are correct. In my opinion, the proper application to Wikipedia is to put constraints on the complexity of individual entries but not on the scope of the overall project.
That makes sense to me. The biggest problem with "non-notable" entries is when they're basically advertised/spammed onto notable entries as "see also" types of links, or even added to running text where they wouldn't normally be important enough to warrant a mention. Orphaned or semi-orphaned non-notable things are relatively benign, so long as they're reasonably verifiable (which a lot of the stuff being deleted isn't).
The biggest remaining problem is search, which could eventually be fixed by a weighted sort of search that trades off closeness of text matches with some sort of measure of notability/popularity, so that "Some guy's theory of physics that was verifiably published in an obscure book in 1933 but which absolutely nobody else agrees with or has even heard of" doesn't come up equally with our "real physics" articles, or even our "famous physics crackpot" articles.
-Mark