Well, fortunately for you, people started asking these questions in the 1970s, so it's already been operationalized and tested.

Here is a (non-APA syntax) reference:

Comparisons of Children's Cognitive Representation of Number: China, France, Japan, Korea, Sweden, …
IT Miura, Y Okamoto, CC Kim, CM Chang, M Steere, M … - International Journal of Behavioral Development, 1994 - jbd.sagepub.com

Cross-national comparisons of mathematics achievement have shown differences in favour of Asian students. This study examined the idea that the superior mathematics performance of students from Japan, Korea, and China may be due, in part, to differences in cognitive representation of number that is affected by numerical language characteristics differentiating Asian and non-Asian language groups. First-graders from the People's Republic of China, Japan, Korea, France, Sweden, and the United States were asked to construct various numbers using Base 10 blocks. Chinese, Japanese, and Korean children showed a preference for using a construction of tens and ones to show numbers; place value appeared to be clearly represented in those constructions. French, Swedish, and US children, in contrast, showed a preference for using a collection of units, suggesting that they represent number as a grouping of counted objects. More Asianlanguage speakers than non-Asian-language speakers were also able to make two correct constructions for each number, which suggests greater flexibility of mental number manipulation. Thus, the unique characteristics of the Asian number language system may facilitate the teaching and learning of mathematics, especially computation, for speakers of those languages.

For what it's worth, there are about a dozen "decent" theories of how kids represent numbers at various stages of their life. The language-connection is one of these theories, but google for language, number, cognitive.

Also, striking differences are observed between speaking and deaf kids.

Please cite examples.

I find it odd syllables have anything to do with the speed of math. I don't speak out the numbers in my head when I'm doing math, I visualize the numeric symbols. I literally see 1+1=2 in my head. I guess the only speed issue would be translating ten thousand five hundred fifty five plus three hundred forty five to 10,555 + 345 in my head, but that doesn't fairly measuring the brain function of calculating the sum.

It seems you can break up the above example into two brain functions. The first is converting higher level symbols such as speech or written out numbers into their root symbol of numbers. The second function would be the mental math calculation.

I assume similar speeds would be measured converting other symbols that have nothing to do with math. However, I do not have the means to scientifically test my above theories so take what I say with a heavy grain of pepper.

Well, this could address a running debate in language and cognition: does the nature of the language affect the nature of your cognition? Based on the language someone speaks, there are certain cognitive operations that are faster.

For example, there is a relationship between the speed someone does mental math and the number of syllables in each number, in the language you think in.