To what extent is all of this a good (solid) objet de recherche?

Just had a conversation with JMD, who is a lot more sceptical about the idea of idiolects. (Si je n'ai pas mal compris) "La langue française n'est pas composé de 60 million idiolectes". I think what she was trying to say is that variation is more down to genres de discours.

I think what I need to do is jump into some data fairly quickly, to have something concrete to talk about.