The variety of tasks they can do and will be asked to do is too wide and dissimilar, it will be very hard to have a transversal measurement, at most we will have area specific consensus that model X or Y is better, it is like saying one person is the best coder at everything, that does not exist.
