Science has both diminishing and accelerating returns. Diminishing because of duplication of effort, accelerating because large projects need lots of people. 1 scientist can't build a LIGO detector in any number of years. Across all the disciplines, linear seems like a good approximation.