Ronald Coase (one of the top 10 economists ever) has a paper about this: firm size is related to factors including communication complexity.

http://en.wikipedia.org/wiki/The_Nature_of_the_Firm http://en.wikipedia.org/wiki/Ronald_Coase

There are big problems I'd like to solve (cheap space launch optimized for LEO satellite constellations, nuclear power), but since firm size for those problems right now is high, I'd rather work on less-interesting problems which can be more readily addressed by startups, at least until technology advances to let <100 person companies work on the big problems too.