I don't think it dismisses the evidence - it agrees with the idea that 10,000 hours might be required. It just states that it takes 10,000 hours for a talented person to become good at a field - that a person untalented in a field won't necessarily become a true expert given enough practise.

To me this seems intuitively obvious - people's brains aren't built alike, and if we're willing to accept that brains at the fringes (e.g. aspergers) are more attuned to certain tasks, it's not too out there to suggest variance within the 'normal' parameters too. Of course, proving it would require a complex test - one that (somehow) sorts people into attuned or unattuned on a given subject to begin with, and then makes them work for 10,000 hours on the subject.

Realistically, though, I suppose this question just isn't all that interesting. A lot of the time, the important part of the 'nature' component is simply what you're interested in. If you happen to have an interest in a subject, you're vastly more likely to put in the 10,000 hours. I'm a good computer programmer because I enjoyed and was good at it from the beginning, so I spent more time at it. I'm a terrible dancer because I spent maybe a hundred hours on it, sucked, and lost interest. Maybe if I'd spent a hundred times longer doing it I'd be a world beater, maybe not - but it doesn't matter, because my initial experience was enough to put me off. I imagine someone for whom it clicked more naturally would be vastly more likely to stay interested.