Another interesting discussion -- as a new visitor to this site, I haven't yet had an opportunity to get tired or frustrated by this topic. I think McGrath has identified some of the key considerations in his very thoughful comments above, but I'd like to add a few additional thoughts.
I appreciate the utility of having clear and consistent definitions to go by (as raised by M.Ted's message), particularly in a scholarly context. But I think we need to be realistic about drawing bright lines around our categories and expecting the rest of the world to conform to them. There will always be outliers to any definition, and there will always be opportunities to split hairs in trying to resolve our differences about what is "in" and what is "out". I think we should recognize that we are not bound by the definitions created by ethno-musicologists and folklorists -- they can use their definitions for their purposes, but we needn't use them for ours. I do not think that our failure to come to complete agreement on the boundaries between musical categories will really impede our ability to make music. In fact, a lot of musicians, in all genres, have created some of their greatest works when they pushed the accepted boundaries. The music will always be more important than the definitions.
Getting back to earlier comments on this topic, I think one additional element that distinguishes "folk" music from other musics is its portability -- or more precisely, the ability to create the music without having access to a large infrastructure. Wagner may have drawn on folk elements (both musical and mythical) in his works, but they were created for large orchestras and opera companies composed of people who had received lifelong instruction in very rigid disciplines -- and that is essentially the only way they are performed, even to this day. Similarly, the music of Brittany Spears (which someone raised) was and is created by corporations, marketing departments, high-tech recording studios with well-trained engineers and session musicians, all with an eye to creating a finished product that will generate the large revenues that are needed to feed the beast. This music cannot exist without the infrastructure that created it, so it cannot really be kept alive by the "folk".
With this consideration in mind, a lot of early rock and roll could be considered folk music, but much of the later rock and roll would fall outside of the definition -- once rock and roll became big business, an infrastructure was created for it, and the music progressively moved away from its roots and reflected the reality of the business that had been created for it. There were periodic attempts to bring the music "back to the people" (the punk wave of the 1970's being an obvious example), but the corporate-based infrastructure was quick to pounce on these, and effectively co-opted them before they really were able to establish themselves as independent forces. In reality, the same dynamic has asserted itself thoughout history in a lot of musical genres, from classical to jazz -- and could soon be a factor in so-called "Celtic" music if the Riverdance phenomenon takes over.
As usual, this is more long-winded than it needed to be (a failing of mine), but I'd be interested in any reactions. Regards. -- Steve