[7777777]

I agree and disagree. :) I think it is silly only if you assume it is incorrect. But, although not from a robust source of data it still could be the correct result. And in that case, I couldn't call it silly. But who knows, the word "innovation" is subjective, so you could probably argue against any dataset, pretty easily.

That being said, I agree that identifying something as an innovation is going to be heavily weighted to past accomplishments. At this point, innovations in theoretical physics are so complex that I can hardly begin to really grasp the importance of discoveries beyond my niche and I myself have a Ph.d. in physics.