Robert Lindsay is interesting, and has a wide breadth of knowledge like many high verbal IQ people (think of Malcolm Gladwell). However, mathematics ability is one of the best indicators of general intelligence, and it’s quite a rare person who can barely even pass Algebra but still claims to possess an full scale IQ > . In other words, I need more evidence.
The best measure of raw fluid mathematical talent is probably the Figure Weights subtest on the WAIS-IV which has a g loading of 0.78. However raw talent only seems to explain 66% to 70% of the variance in various cognitive performance, suggesting talent only correlates 0.82 with performance. If we assume that the correlation between g and math performance is caused simply by their shared correlation with math talent, then the correlation between g and math performance is simply the product of these two correlations: 0.78*0.82=0.64. Commenter misdreavus strongly objects to this type of reasoning, but it was endorsed by none other than the great Arthur Jensen.
A 0.64 correlation means that that someone with an IQ of 140 on a hypothetically perfect measure of g (2.66 SD above the mean) would on average be 2.66 SD(0.64)=1.7 SD above the mean in mathematical competence. The standard error of the estimate would be 0.77 SD so you could say with 95% confidence that a 140 IQ would have math skills between +0.16 SD (55 percentile of the general population) and +3.24 SD (99.95 percentile). Robert Lindsay estimates his math skills to be around the 70 percentile, putting him within this statistically expected range.
[UPDATE, JUNE 2, 2015: The end of this post has been edited based on feedback in the comment section]