While Wikipedia strives to avoid self-referentiality, sometimes it cannot avoid self-referentiality. Kottke’s example above is one that’s likely to stay, and for good reason. But another comes to mind, although you have to dial back the clock a few years:
Yes, the Wikipedia article “Original research” once carried a warning asserting that it contained original research (a big no-no on Wikipedia). Today, “Original research” is merely a heading within the larger article “Research”, which is probably as it should be.