I've seen very hard problems solved in a way that creates a lot of cognitive load for other teams, and I've seen hard problems solved in a way that significantly reduces the cognitive load of other teams.

Consistently adding avoidable complexity has a very real impact on velocity.

Good senior/staff+ engineers are able to solve hard problems.

But what sets apart some of the best I've worked with is that they are able to solve hard problems by introducing the least amount of complexity.

What I really want to emphasise, is that the gap between good and great can be huge.