I could also call McColl an English folk singer, songwriter, labour activist, actor, poet, playwright and record producer. None of which gives him any particular qualification to define what is or is not folk.

He didn't. The 1954 definition was mainly constructed by Maud Karpeles to reflect a partial consensus among a bunch of ethnomusicologists of the period. If MacColl went along with it, it was simply because it was the clearest thinking on the subject you could find at the time.

Cowdery's paper is a nice summary of what happened.