analogous to fitness in evolution
Agreed but for a different reason -- both are subject to high degrees of random pruning. The end product of natural selection is often not the "most fit" of all the variants ever born, but simply the most fit of those not culled by random disasters and pruning. Ditto for innovation impact, where we can clearly see that new innovations that become the standard in a society are often not the 'best' by what we now think are natural metrics of fit [network effects, social norms, many other adoption ratchets dictate what's possible]