The problem is that you're portraying this as some sort of law of finite biological resources or something, rather than a probabilistic effect of breeding small

The problem is that you're portraying this as some sort of law of finite biological resources or something, rather than a probabilistic effect of breeding small populations. Most genotypes or "random mutations" are totally neutral for an organism, but let's say that of the notable ones, it's 50/50 whether they will be "good" or "bad". If we find the "good" smart parrot, it's likely they have some other "bad" stuff we can't see or just don't care about. When you keep inbreeding, that other bad stuff will be fixed along with the intelligence. This is purely a limitation of our knowledge of the genome and lack of other inputs. You can out-cross to reduce the bad traits, but you risk losing some of the intelligence too. Theoretically there is another bird out there that solves the "bad" trait without affecting intelligence, we just don't know which it is. Nothing stops you from having a win-win-win. You could randomly find a bird that is super smart with rainbow feathers and the voice of an angel... It's just incredibly unlikely to roll all those attributes in one individual and keep them through multiple generations. As to why wild populations haven't naturally evolved to be smarter, that is an entirely separate discussion about countless competing selection pressures and how genes spread through larger populations.