I don't know enough about neurology to know whether that's possible...
I believe it to be - I would have to do more research to give you an exact example of improvement, but I do know of an example of the reverse. A zebrafish mutation named
space cadet causes the loss of a neuronal connection. When normal fish are poked, they respond by swimming away quickly; when the
space cadet fish are poked, they turn around and swim into the threat. (Obviously this mutant would be selected against in the wild...)
if it is, whether it would be long term and as to what effect it would have.
It would be a genetic change altering neuronal development, so it would be permanent for the individuals with the mutation.
Often people think in terms of what they view as "simple" traits, such as pigment/camouflage, to discuss the effects of mutation/selection. As you stated above, you accept that mutation producing a differently colored fish could adversely effect its survival.
However, though color may seem "simple" to us, it is
incredibly complex, and very little is known especially about the production of pigment patterns. "How does the zebra get it stripes?" remains a very unanswered question. Indeed, the connectivity of the butterfly neural complex may be less complex than the "connectivity" of the countless cells interacting to produce the pattern of a butterfly wing.
What I'm saying is that we shouldn't consider neurology as necessarily more complex than pigment and pattern - such things are likely our own misconceptions rather than the reality of the situation. Likewise the effects of mutation and selecion on a variety of things, especially neurology, can effect survival rates in regards to predators.
I'll be interested to hear your response - I'm still wondering if you have an idea of a "complexity barrier" that mutation/selection can not surpass...