>he worked on problems that interested him Yes, one can only engage with a problem that has personal relevance. Mostly people aren't engaged. They are notional

>he worked on problems that interested him Yes, one can only engage with a problem that has personal relevance. Mostly people aren't engaged. They are notionally working on 'important problem X', but their real problem is how to please an authority figure or how to further a career. Which aren't interesting. A study of history shows that objective progress is possible, so it follows that it must come about by lots of Feynman-like people working on their own apparently unimportant problems. This can be explained by the way reality and our knowledge of it are structured. Solutions to problems typically have 'reach'. Thus in order to solve parochial problems people typically create fundamental theories which automatically apply to all sorts of other problems, including important ones. The Beginning of Infinity by David Deutsch is a fascinating read which has a lot to say about this sort of thing.