GOT

Inspired by TOP, I've run a few geometric complexity weighted
optimizations; if you adhere to the philosophy that lower-complexity
ratios should be weighted more, the results seem reasonable.