Peter:
We all let our designs evolve. But, we try to complete the process in the dirt. By the time we decide to install the irrigation and plant the grass on them, our clients [the actual course owners] hope that the evolution by construction is done for a while, because it's vastly more disruptive (and expensive) after the fact.
A lot of the criticism for architects and courses that go back and change stuff after the first year or two, is simply that it's an indicator that maybe the lead architect wasn't there enough, or wasn't paying close enough attention at the time, and only figured it out after the fact when there was a clamor to solve some problems. Sure, there are sometimes playability issues that you didn't foresee ... but it's the job of the architect to try and foresee them, and we have to accept the criticism when we didn't.
There are certainly parts of a golf course that can evolve after the fact. Grassing lines can be changed, and the density of rough can be revisited. A new tee can be added here or there, relatively inexpensively. Bunkers can be added or taken away without much trouble. But, rebuilding greens is not a normal evolution.