Given the other metric this week, just for fun, I placed extra arms in every available spot in my original solution, throwing the cost tertiary to the wind. But what would happen if someone really optimized for amarms as a secondary? Maybe one day we'll find out...