I suspect we'd have an innate/growth model issue here and would have to be careful to define our terms and question carefully. For example, my DS is doing well with piano at present and gets described as "talented", but I suspect that, like me, he will turn out to feel that he "really" has only very moderate musical talent - enough to enjoy music socially but not enough to make a career out of it, for example. In the early stages of learning an instrument, a combination of high general intelligence (problem solving ability, applied to decoding music and producing the sound your teacher asks for!) and greater-than-average willingness to work hard (and/or a parent who is willing to enforce it) will let you advance fast, for a while. However, it won't necessarily let you keep up for ever with the person who had typical general intelligence and age-typical application skills but who seemed to know instinctively what sound to produce and how.
I guess what I'm saying is that if you looked at people with say under 5 years of instrument learning, I think you'd find a strong correlation between academic and musical performance levels; however, I think if you looked at the same people 5 years later, you'd find the correlation much weaker.
(All scare quotes in this post are deliberate :-)
