Agreed; someone who says "engineers produce their best work 2-3 years into a role" is a manager who has created an environment where an engineer's best work is produced 2-3 years into a role. This isn't a recognition of some universal truth; its an observation of the environment they had a hand in accidentally engineering.
It is similarly easy to envision an environment engineered to encourage engineers to produce their best work 3 months into a role; Give new engineers one really exciting, well defined, small greenfield project, after that move them into a legacy maintenance role, never give bonuses, raises, etc. I know I'd get burned out pretty quick; my best work would be very early on.
It is harder to envision a systemic change to encourage engineers to always have produced their best work yesterday. I'm not suggesting its easy; by far, the easiest and most common course of action is exactly the environment the parent commentator engineered, where engineers peak a couple years in. But, don't be fooled; the engineering division absolutely suffers in this environment, and it is possible to do better.
It is similarly easy to envision an environment engineered to encourage engineers to produce their best work 3 months into a role; Give new engineers one really exciting, well defined, small greenfield project, after that move them into a legacy maintenance role, never give bonuses, raises, etc. I know I'd get burned out pretty quick; my best work would be very early on.
It is harder to envision a systemic change to encourage engineers to always have produced their best work yesterday. I'm not suggesting its easy; by far, the easiest and most common course of action is exactly the environment the parent commentator engineered, where engineers peak a couple years in. But, don't be fooled; the engineering division absolutely suffers in this environment, and it is possible to do better.