See but we know that things like nutrition, toxins like lead, education, wealth, family structure, infection rates, and other factors that also differ between nations affect intelligence. That's not even getting into Differential item function, the use of IRT vs fixed testing or other statistical and psychometric factors that may impact the measurements.

You can't just compare two samples without controlling for confounding differences.

Science can't even fully explain the Flynn effect, in which industrial nationals have had IQs going up 3 points per decade on average within the country, and people are trying to use IQ to talk about differences between sample populations.