In looking back, revelations that have improved the way I work were always the result of having tested something inferior before that. Rarely do we jump into the game (whatever game it is) and hit a crescendo from the outset. People who can do that are the .0001 percent, and we should probably be more forgiving with ourselves when it comes finding better, more efficient ways of doing the work we do.
Good is how we get closer to perfect, and we won't know what should be tweaked unless we run experiments. Impact, productivity, and enjoyment are the direct result of solving, launching, & iterating.
Let's not be overly precious about the things we're building.
Let's not let perfect get in the way of good... because perfect isn't the goal. Better, is.