I think the notion of local minima fits into this. It might be optimal for male birds of some species to have very colorful flair to attract females. Within tha

I think the notion of local minima fits into this. It might be optimal for male birds of some species to have very colorful flair to attract females. Within that species of birds, these males with the more colorful/visible/expensive plumage will be selected for. This will remain true on shorter evolutionary time scales. It will work for a while. However, if the males of the species develop flair that becomes too attractive to predators or makes them incapable of outrunning said predators, and most males get eaten, natural selection will ensure that things balance themselves out. That species of birds could go extinct, or it could be that there are very few colorful males left, and the females who are less selective win out because reproducing with the males that are left is better than not reproducing at all. Evolution is non-stationary. What's making an individual fit depends on its current environment here and now: the food it can find, what predators are out there, things that make it more attractive to mates. However, predators, food, weather and mates change over time. In other words, what's being optimized for is constantly changing, there are local dynamics and some amount of "noise" that occurs, possibly resulting in silly looking animals like peacocks happening once in a while.