IanC,

Very interesting. Really enjoyed your posts.

Some remarks:
1.
It seems that the one of the most significant problems with certain kinds of "internal evidence" is the rather fluid nature of the "details" found in traditional songs. A classic example is Cas Wallin of the US who sings a version of Matty Groves which names the tattletale footy page as Robert Ford, killer of Jesse James.

Mondegreen-ization also causes such fluidity.

2.
I was also struck by the following in your discussion of "The Black Velvet Band":
"These broadsides are almost unsingable in their original format, and there is little chance that they originally derived from an existing traditional song..."

It looks to me as if there is a significant presupposition about "singability" lurking behind this claim. What evidence do you have that your notion of "singability" coincides with that of the 19th century target market for such broadsides? I ask because I agree with Q that "A real problem for many of us is trying to match our thinking to that of people of an older time period when we are trying to evaluate a possible piece of evidence."

3.
Finally, a modest acquaintance with "historical Jesus" scholarship leads me to feel that a consideration of the structure and form of a song would be a necessary component in determining age.

From my own experience, some ballads sound to me like they were intended to be passed on orally and some sound like the were intended to be read. I would assume the former are older.