This is true, but the saying "time is money is also true". If you are in a situation where you can throw a thousand dollars at something that will save you hal

This is true, but the saying "time is money is also true". If you are in a situation where you can throw a thousand dollars at something that will save you half an hour in turn around for each build, then that would probably be a wise choice. However once things have solidified, it would then make sense to look at optimization. You definitely have to be smart about your time, especially if you value it. Otherwise you can find yourself spending a week to optimize, plus test to gain very little performance. I'm a strong believer in throw hardware at a problem if you can in the short term to buy you valuable time elsewhere.