[raganwald]

I would say that any number of individuals "after the fact" are kind of weak as evidence. The very best empirical science is making a prediction of some new thing we've never seen before, and then testing the prediction.

So we don't dig up a bunch of bones and look for evidence that dinosaurs evolved into birds, we predict that if we dig, we will find something that looks a little like a dinosaur and a little like a bird, and then we dig.