After reading these comments I think I might have understated my original point - to wit, it's easy to overdo the fun parts, and you need to measure what you are spending on the fun parts, because the true cost might be that of future fun parts down the road (more time to work on cool new technology a year from now) or a lot of money.