I read as much of this as I could stand, then skipped to the bottom to add my two cents' worth. Some of you already know my position on this topic ~ so feel free to skip this, if so inclined..

As a (retired) street performer, acoustic guitar and vocal, I always made it a point to construct my repertoire using selections (A) that I know and can perform credibly and (B) that people, in general, like and recognize ~ and might, for example, be motivated to sing (i.e., "sing-along") on a long bus ride or around a campfire.

Such a repertoire includes quite a few "popular" songs written and recorded in recent decades, and does NOT include songs that are without-a-doubt "traditional" but which are obscure and unlikely to arouse audience interest.

Does this define "folk"? For many (e.g., Cecil Sharp), obviously not.

But for me, such songs DO comprise the "folk culture" of our 21st-century, electronically-linked, world-wide community of musicians and listeners ~ in very much the same way that a given selection of traditional songs may have defined the folk culture of a 19th century fishing village or an 18th-century farming community.

I would consider, for example, that a number of Beatles songs have attained "folk" status under my definition ~ but that a precious few compositions by "folkie" singer-songwriters have achieved anywhere near the kind of widespread recognition that earns a place in our contemporary canon of "people's music."

Another thought: if you consider blues to be folk music (even if you restrict your definition to acoustically-played blues songs), it is almost impossible to exclude many songs with known composers.

I will draw the line somewhere though; to me, the Coke commercial about teaching the world to sing is NOT a folk song. But maybe that's only because I don't like it very much at all...