This should become clearer with a full detailed study of all the Irish versions. One also can't dismiss the well documented sexual connotations of 'well below the valley' though one could hardly think this was deliberate. If it derives from anything I would say it comes closest to the Percy version and I repeat what I have suggested before, the idea of such a ballad surviving for 3 centuries with no intervening appearance is highly unlikely but not impossible.

The Scottish 20 additions could easily have come direct from Percy as plenty of Scottish antiquarians were corresponding with him.

I do agree the evidence is beginning to show that it is unlikely the Irish version came from Scotland.