I knew I'd have more to add...

"Did he at the same time filter out any obviously Scottish or Irish songs he came across?"

There's no evidence of that at all. As I've said, he spread his net much more widely than purely English material, and he wrote that many of the ballads were often to be found in Scots collections like Kinloch's. One thing to bear in mind is that many of the ballads we associate with Scotland, like 'The Dowie Dens of Yarrow', 'Andrew Lammie' or the ballads of Border skirmishes, seem not have spread much to the Appalachians at all - perhaps some of them post-date the migration? And, where Sharp and other collectors in the mountains found ballads popular in both England and Scotland ('Two Sisters', 'Geordie', etc) the variants looked more like the English versions: 'Binnorie' variants of 'Two Sisters' are virtually unknown in the mountains, and most of them don't include magical harps or fiddles.

As to the singers themselves, Sharp wrote that the people in one area he visited were supposed to be 'Scotch-Irish', but that he couldn't discern any specifically Scots or Irish characteristics. He did however refer to one singer in another place as 'a tall Scotchman', and noted that a woman singer he and Maud became friendly with was 'Irish cum French cum Indian'. So he wasn't rigid in insisting they were all pure-bred English, although he did generalise in those terms in his introduction to EFSSA.