Thank you for those, Folkiedave. The Bearman paper in particular provides some very useful hard data on what C# did actually collect and what he did with it - previous contributors to this thread please read!

In an appraisal of C. J. Bearman's work here Mike Yates makes rather well the point I've been trying to get at:

"Does all this nit-picking really matter? Well, yes it does. Because if our foundations are based on false assumptions, then the whole subsequent body of folksong and folklore studies is liable to come tumbling down around us."