John G.,

I agree from your evidence that the Alabama scenario is viable. I'm just saying that showing how "Margaret" could change to "Polly Ann" doesn't really add anything, because it's all guesswork. Any name could be transformed into any other by a series of similar assumptions and logical leaps. It's not in itself evidence, and neither is it logically derived from evidence, because as I said it is logically circular--you need to know what conclusion you are coming to before you start, or else you will never get there.

Your own statement that "When you come across something that is both unfamiliar and complex, or seems out of place, such as 'Mary Magdalene,' I think you should automatically give it great credence as a possible 'original,' or a relative of the 'original,' and try to check it out further" is not in fact standard practice in folklore or any other discipline. To give an example from my own area of the country, if I came across an informant who claimed that the Jersey Devil was a deformed boy born to Migdaloosa McChuzzleford, I would not automatically assume that this was any more likely than the usual "Mother Leeds" or "Mrs. Shourds," nor I think would any other folklorist.   

Your contention that Maggie is a common nickname for Margaret is true, but there is no Maggie in any of the songs either. I may have missed something, but I don't think you've even established that Margaret Dabney was known as Maggie D.

I'm not saying you shouldn't "check it out further," but so far you haven't really checked it out, you've just shoehorned the name into your pre-existing hypothesis of Henry Dabney. Once again, I don't think this particularly hurts your argument overall, it just doesn't help it much.