Fakesong promises far more than it delivers, to my mind, and then claims to have delivered it anyway. If somebody were to put in the hours and identify how many of (say) the Child ballads were relatively recent parlour compositions & how many probably had genuine popular roots, that would be really interesting. Harker didn't take those pains, possibly because he was committed to discrediting the idea that any of the canonical ballads *could* have had popular roots - or rather the idea that we could trust any of the claims made to that effect. And fair enough, on one level - better scepticism than credulity - but if you use too high a concentration you end up dissolving the subject altogether.