Thanks for all that Brian.

I tend to think about 'sampling bias' because I have had to in other contexts. So far as rescuing a repertoire 'for use' is concerned it probably isn't too important if a tune or text from popular music of previous times gets included (and it might fall within the 1954 definition). Missing local compositions that sounded like earlier popular music could be a loss. Also I am not sure if some collectors also had a 'quality' criteria - I often enjoy a bit of doggerel in the right context.

As for the tunes, my suspicions are often raised. Walter Pardon recognising old tunes because they end on the draw caught my attention because I knew Sharp's figure of about 2/3 major. There is a tendency for the modal tunes to get talked about more (including by Sharp) because they are uncommon in art music and for them to be favoured by players because they are cool and different from a lot of what we hear. In the few dozen dance tunes from mainland Europe that I play (because others locally do) there is a suspicious predominance of those in a minor key.