Turtle Old Man's question is a good one. Look at JJN's descriptions of the singers in the Ballad Book--most of them do not give the sort of information one would need to find the alleged singer, his or her descendents or neighbors, or anything of the sort. In addition, the Ballad Book was published only decades after the songs were supposedly collected, and most of the singers would have been dead anyhow. The sort of info we find in the standard collections--Collected from John Dinkydoodle, age 58, East Grumpus, Kentucky, June 14, 1926--is notably lacking. Instead, we get vignettes about colorful characters with no surnames. If you ask someone where he got a song and the reply is something like: She was an old woman with a lively demeanor, surrounded by three generations of descendents, and she sang with a voice still clear but so soft that her words were sometimes hard to make out. Her homespun dress was getting threadbare but was freshly washed.--And so on, you may begin to wonder whether this is a real person or an invention.

Still, I wonder as I wander whether JJN's collection does not include songs just a bit retouched, still close to what he heard from some mountain man or woman. He liked to point out that he had spent years travelling about in the mountains, in contrast to the comparatively short but immensely productive visit of C#, and that he looked for unusual versions. That does not seem unbelievable. If he had been more up front about his outright compositions and his revisions of oral versions he actually heard, and had supplied usable information about his sources, we might be able to cite part of his collections with a good conscience. I don't find his "Little Mattie Groves" unbelievable as a traditional song, for example; but given the circumstances I wouldn't cite it among the traditional melodies.
If I recall correctly, Bronson included his versions in the first volume of his magisterial collection of melodies of the Child Ballads, and then eschewed them in the later three volumes, and since then it has been pretty generally accepted that his versions are not reliably traditional. If some of his songs are sufficiently authentic to be included in the scholar's corpus, he brought it on himself that the entire corpus is regarded as suspect.   

I find many of his songs quite beautiful (and also don't have the negative reaction to his singing that some experience); I certainly don't have a negative opinion of him in general. But I have to say that if no one checked up subsequently on the singers he supposedly got the ballads from, he made it pretty well impossible to do so, and that in itself does not inspire confidence in his accounts of the singers, colorful and sympathetic as they often are.

Stephen