I don't see how your contention really fits - physics Olympiad types are by definition not general.

For what it's worth, when I was in that position I found some of the Landau and Lifshitz stuff at least as insightful. I'm all for Feynman as a supplement, by the way.