If forced to choose, i'd probably pick the feedback loop theory as well though.
It also seems possible that there are certain aspects of language and culture that support math achievement. Gladwell, in _Outliers_ is after this category of explanation.
Personally, I tend to like cultural explanations more than raw talent explanations because I like to think that everyone is capable of everything (eventually, if they work hard enough). I certainly like to act like everyone is capable of everything and it helps to have supporting theories.
Gladwell’s theory goes something like this: First, the counting systems tend to be more consistent, less idiosyncratic. This means kids learn them better earlier and build confidence early. This early start builds on itself — a point he comes back to over and over in his book. Second, the numbers tend to be faster to say. This means you can remember more digits — quite a few more. This has the same impact on confidence. Third — and this is his main deal — rice farming yields to careful, persistent effort much like a math problem. This is less true of western farming. All the western farming sayings go something like “there’s no accounting for the weather. sometimes your just eff’d!”. All the Asian sayings go something like “work hard and your family will prosper for generations.” He says this gives Asians a leg up in math because math yields to the same kind of persistent effort.
Also — someone below diss’d on Gladwell calling him a “pop” psychologist. I don't think he is not a psychologist, though I guess he is pretty pop. Still, what he does, while researched, is not science, that’s for sure.
I think explanations like Gladwell's are valuable even if they are not true because they at least show us how nuanced and twisty the truth might be. That keeps us looking for explanations that are maybe not so simple. That's good because as a species we're stupid enough without the crippling detriment of thinking we already understand what we do not.
EDIT: very minor for clarity.