It's about trade-off. You improve performance at the cost of something else you're not working on. Without the context, it's hard to tell if that cost was justi

It's about trade-off. You improve performance at the cost of something else you're not working on. Without the context, it's hard to tell if that cost was justified. I.e. maybe there were 20 critical bugs and the team made the call that perf. was good-enough and it was better to stabilize the software. Most often than not, you can always improve performance and drawing the line is hard.