A few pointers to work on little-known, -recognized, and/or -understood aspects of whatever we're doing when we're not necessarily reading but -- for lack of a better way to say it -- just generally doing things with stories that whose endings might come sooner or later than we want them to (but never, it seems, right on time):
Carl W. Scarborough, Godine's book designer, on The Haptics of Reading. File under general problematique of What you read affects how you read it, AKA all reading is not the same activity (which opens the possibility that some it may not in fact be reading at all). A snippet: "There is a dramatic difference between the sensations inherent in reading a novella—necessarily a small, intimate book—and studying an overscaled art book, where the size of the illustrations plays an important role in the satisfaction we find in reading. [...] Another aspect of this haptic issue is the question of suitability of materials. I have beautiful twentieth-century books printed letterpress on exquisite eighteenth-century papers. Reading them is a special pleasure, but it is very different from reading an art monograph where a satiny, coated white sheet makes the illustrations leap off the page."
Rubin, D.C. (1995). Memory in oral traditions: The cognitive psychology of epic, ballads, and counting-out rhymes. New York: Oxford University Press. May suggest an answer to the question, why are so many epics so similar in outline, but not in detail? From OUP's blurbage: "Focusing in particular on their three major forms of organization--theme, imagery, and sound pattern--Rubin proposes a model of recall, and uses it to uncover the mechanisms of memory that underlie genres such as counting-out rhymes, ballads, and epics. The book concludes with an engaging discussion of how conversions from oral to written communication modes can predict how cutting-edge computer technologies will affect the conventions of future transmissions."
Rubin, Ciobanu, et. al., "Children's Memory for Counting-Rhymes," Psychonomic Bulletin & Review, 1997: A suggestive study of children's rhymes like "eenie meeny miney mo," which have to end a certain way, in a certain time-frame, but still support intra-rhyme variations. The rhymes are like games in their playful purposiveness, in being directed toward a goal. But they are also "oral traditions" & so share affinities with more exalted traditional oral forms, like epic.
A post at Obscene Jester arguing that "Lost" is pernicious narrative dreck because it flatters the viewer's (idea of) her own cleverness. If this person is correct, then people who like the show must be getting their satisfactions from somewhere besides the neat narrative closure we are all conditioned to expect from popular media forms. Sometimes the complaint "it goes on too long" (and its other variant, "this story is plotless," or even occasionally "it is too much like a hypertext") is shorthand for "this media makes me feel like a moron [so therefore I hate it]." Commandment Numero Uno: Thou shalt not trample upon thy audience's self-regard!
Recent updates on these topics from Mark, J Nathan Matias & Jill.