This does not apply to Bunting's method, but the most astonishing key signature I've seen in a folk transcription is G sharp minor, in a Scottish MS from the early 19th century.

The tune is very similar to fiddle tunes in G minor, and that's the key you find it in from the 18th century sources.

My explanation: the transcriber was playing the piano alongside a fiddler and the piano had drifted a semitone flat. He wrote it down the way the piano was playing it rather than how the fiddler saw it.

Wishful thinking from overeducated transcribers is alive and well. Look at any of those recent Scottish tunebooks with Celtic knotwork everywhere - all done by fiddlers, and pipe tunes get their G's sharpened systematically wherever the editor saw a V-I cadence coming. This is the same failure of taste Dow was making 200 years earlier.