I think your distinction is pretty arbitrary; in what way does a mathematician not attempt to reason about the world in a systematic fashion and according to empiricism and logic? The only real difference over time has been in our ability to determine valid logic, our rigor in what empirical evidence we'll accept, and our sophistication in how we systematically go about this process. Early philosophers were attempting each of these things to various degrees, they just weren't as good as we are now at them. Really defining the "first scientist" is like defining the "first mammal"; the category is fuzzy and doesn't really mean anything at that level of granularity.
I suspect Keynes’s desire to discredit Newton was motivated more by the former’s irrational, perhaps as the first postmodernist, impulse to deny the value of gold.
The evidence is that if you go to the top philosophers these days then they will be unaware of the right answer to the first question in studying philosophy: what is the definition of philosophy? Without having learned it, people have been going around trying to teach what they are unaware of.
Even in modern science, there is a lot of ambiguity about what scientific method 'permits'. For example, many physicists would scoff at the idea of String Theory being called real physics.
What happens, if, in some 500 years we find a definitive way to simulate experiments? I can see any physical science not validated by such a simulation being dismissed.
"Scientist" is a modern term and I would grant historic figures only that label, if they (disregarding if their conclusions were right or wrong) tried to reason about the world in a systematic fashion and according to empiricism and logic.