1. At Crooked Timber, Henry's post on an article I can't read is now spinning off into a more general discussion of who cites what when and why. Since I just spent this afternoon looking, once again, at Book Two's footnotes, and noting that I probably should cite a few more people here and there (although there's nothing I can do about this at present...), this is a question near and dear to my heart. The greatest amount of juggling, I think, occurs in what one of Henry's commenters calls "string citations"--those long, long, long lists of everyone who has trod the stony paths of whatever happens to be your current argument. If, for example, I have two people who have already made claim X, should I cite a third who says the exact same thing? Four people? What about five? Having cited five, what's the likelihood that the editor will then ask me to reduce the list to...one?
(And, of course, sometimes you just forget to include someone in your string. Until a peer reviewer says, "um, where's so-and-so?")
Another poster hints at the problem of articles published in the "wrong" journal, but this is an issue that goes beyond academic politics. If the article is in a "strange" place (e.g., a historical journal that rarely publishes on literary topics), then nobody will expect to look for it there; more to the point, it may not be indexed by the usual suspects. For example, Kevin L. Morris' articles on Catholic and anti-Catholic literature, published in Recusant History, aren't indexed in MLA. Nor, for that matter, does MLA magically index everything, as I know all too well: look up "Grace Aguilar" in the MLA index, and my own article fails to appear. (This was even more of a problem in the early phases of digitizing the MLA bibliography, when you had to check the early hardcopy volumes after searching the CD.) Similarly, a book from the "wrong" publisher may be overlooked for any number of innocent reasons, ranging from lack of publicity to, again, any given scholar's assumption that nobody in their right mind would publish Book X with Publisher Y. (Or, contrariwise, that Publisher Y would normally keep its distance from Book X.) From a scholarly POV, one of the handiest things about Google Books, Amazon's Search Inside the Book, Project Muse, and JSTOR is that you can search for content that escapes the clutches of the main bibliographies.
2. D. G. Myers' call for literary history elicited several queries from one commenter, R.T., including this one: "Do some (or even many) literary scholars avoid new territory because there is so little in the way of secondary sources which they can use (either through contrast or comparison) to further their own arguments?" To a great extent, though, the lack of secondary sources is an illusion. The world is not exactly overflowing with articles or books on religious historical fiction, although there are a few, but writing about such fiction requires you to read a lot of other things: history of religion; theory of the historical novel; studies of nineteenth-century historiography; studies of canonical novelists writing about religious topics; studies of the period your authors are writing about; etc. The same would be true if I were writing about twenty-first century apocalyptic fiction.
I should also note that a great deal of pioneering work on non-canonical authors was done in the first half of the twentieth century, both in dissertations and otherwise. (Women writers, Victorian periodicals, genre studies...) Writing about non-canonical authors used to be a conventional approach to a dissertation, thanks to the Germanic model of doctoral research--what one of my professors at UCI called the "Turks in the English Renaissance" approach to dissertating (i.e., find every instance of a Turk in the English Renaissance, then explicate function of same).