exactly... inflexibility to change an old process because of 1% or less of the use cases could be justified or not. If a new process makes 99% of things take 10% of the time and 1% of things take 1000% of the time, you still win. I would argue that edge cases should rarely be a reason to hold up improving a process.