There are times when optimisation is a good idea. For example:
When the technology involved is mature. With a rapidly changing technology, process optimisation may not keep pace with technological evolution.
When the environment is stable. It is easier to optimise a structure for a prevailing wind than for blustery conditions.
When customer behaviour is constant. If customer demand is broadly unchanging, then we can optimise around how we carry on giving them the same thing.
When you have good feedback. This is critical. Without good feedback from the system you are operating in, you don’t know if what you are putting into that system is meeting your aims. And you can’t see if the system conditions are changing.
When there are no disruptors. These disruptors could be technological. Or they could be a group of engineers (or other humans) with an approach that is changing things up. It is too late for optimisation when no one needs what you are offering.
In short, optimisation is good when the conditions are steady.
But if our operating environment is changing, then we need to dedicate at least some of our resources to asking, do we need a different approach?