Seems to me there is a difference between what you can get away with singing in many gatherings of those who consider themselves 'folk aficionados', and what was sung by those in an earlier day who didn't KNOW they were doing 'folk'.

I see several posts in which those who espouse a pretty broad definition also note the limits beyond which they would NOT consider certain music to be folk. Perhaps if we take all the definitions and combine them, almost nothing would be excluded.

I have tried for years to make the point that if you are going to use the word, it has to have a narrow enough meaning to be useful....otherwise, the category ends up just being "music"..(or "music *I* like").

   I have proposed more than once the concept of taking a bunch of music/songs and submitting them to some sort of analysis, such as greg stephens notes above..."There are three questions to ask: how was it made, what does it sound like, and what is it used for? Any one of the answers can make it folk.?"
   There could even be several more criteria, depending on whether you refine those 3 ...'age', 'anon', 'method of transmission'...etc..but the idea is there.
Given this concept, I'd guess that most of the songs in the Digital Tradition database would pass the test...that is, that most 'folk aficionados' would agree that they are 'folk/trad'.

Now...take a bunch of music that is not in the DT, but is widely heard done by people with guitars and much of which is written BY self-proclaimed folk aficionados, and you start to get some differences of opinion. You begin to get those "grey areas".
   At this point, it becomes a matter of what you can get away with. You know that there are some places and some groups where you will cause an uproar if you try to do music outside certain bounds.
So...are we just interested in deciding what grey areas will be tolerated in lots of places, or are we concerned with abstracting from all this a reasonably concise linguistic definition, like the 1954 attempt, which attempts to explain why those difference exist and keep the term narrow enough to have real usefulness?

We will go on singing the same songs...songs we like to sing: but maybe we can at least be more aware of their place in the hierarchy, even as we participate in the eternal process of 'processing' them INTO 'folk'. ...(I know *I* sing many songs which would not fit under the narrowest type of 'folk/trad', but I also know that I am deviating, and there are places & circumstances when I would avoid those from the grey area.)

So,,,is that convoluted enough?