The article is nonsense that exists in the realm of fantasy. The reality is that there are two opposite but equal dangers: smart people tend to overestimate average intelligence, but they are just as likely to underestimate it due to hubris and a desire to feel special.
There are lots of red flags in the research he's presenting.
The first, and most common one for research of this type, is not considering different cultural values and conditions. For example, East Asian cultures value doing well in testing vastly more than any other culture. This itself would skew the data, as two equally intelligent individuals are likely to give different test results if one is vastly more motivated.
Indeed, it turns out that financial motivation
causes a significant increase in IQ. Notable, this effect is stronger on people with lower IQs - so on average, a person with a tested IQ of 85 suddenly shows an IQ of 100 when money is involved, but a person with a starting IQ of 110 only moves to about 120 on average.
This means that, if two countries truly do have a discrepancy in IQ, a lack of extrinsic motivation (which is definitely the case in the tests the article is talking about) will exaggerate the IQ gap between the two countries.
Indeed, someone in the comments of the article sort of brings up this issue: average British results for the math GCSE (an important college entrance examination in the UK) greatly contradict the average British teens math ability per the research he is citing. The difference between the research he is using and the GCSE is that one involves significant motivation and the other does not.
The other is the problem of less than 5% of individuals answering a multiple choice question correctly, especially a question that has been translated. Even if the people involved are in fact bad at math, such consistently wrong answers to a translated question to me suggests a problem with the translation. For example, he states that only 2% of Jordanians, Colombians, and Indonesians successfully passed Level 4. Yet 8% of the population
guessing would result in at least that level of success on a four possible answer MPC question. If the people involved were truly that uninformed, I would expect to see a higher percentage of guessing than that.
This is before you get into the fact that
black people are more likely to guess an answer to a question on a test than white people, and men are more likely to guess than woman.
Honestly, I don't believe in intelligence research that doesn't result in an actionable plan.
For example, if I do market research, and come up with several plans targeted at different demographics, and those marketing plans success, that is strong evidence that my market research is good.
I don't really think there is any equivalent in the area of comparing levels of intelligence between countries. You can't really create a plan that successfully achieves a 'goal' that shows that your research is good. Instead such research generally comes up with stats, and without those stats being used to predict future outcomes, its very hard to actually know if they are any good, especially since modern scientific research is so poorly protected against researcher bias.
There are a myriad of other issues in this article, I'm sure. Probably plenty I'm not even aware of.