The 250:1 substitution doesn't work because breakthrough AI research follows power law distributions where top researchers produce orders of magnitude more value through novel insights that can't be replicated by simply adding more average contributors.