I wonder how much of those things he actually figured out in his 20s and just did not publish, versus thinking about them without drawing many conclusions or making significant progress until he actually published 20 years later?