For purposes of this discussion, I'm not trying to identify the causes of the differences between the sub-populations. My point is that if you are talking about the quality of the educational system--which is what this discussion is about--you need to compare apples with apples between countries. And to do that, you need to account for the fact that the U.S. sub-populations aren't equally situated.
For example, Asian Americans outscore Japanese, Korean, and Taiwanese students in PISA, including math. That's not a cultural difference. That's because Asian Americans aren't a random sample of Asians. The vast majority are within one generation of a very tough selection filter that screens for high skill, high intelligence, and high motivation. If the point is comparing schools, it doesn't make sense to include Asian Americans in the average.
> I get that you're more familar with US society, but this is a thing basically everywhere.
It's not a thing in the east Asian countries that top the educational charts, like Japan and Korea. Poor Japanese and Koreans still belong to the majority ethnic group, speak the national language at home, etc.
Say you transplanted Japanese or Korean schools into one of the many majority-Hispanic school districts in the U.S. where most of the kids are children of low-skill, non-English-speaking immigrants (often illegal immigrants). Would those Japanese or Korean schools have higher test scores than the American ones? I suspect they'd actually be worse, because they'd be totally unequipped to deal with a large student population from a non-native language background.
My wife's aunt's kids go to a school in a more rural part of Oregon. Many of the kids are children of agricultural workers. Many of these kids don't even speak Spanish at home. They speak one of dozens of different indigenous Latin American languages. Japanese and Korean schools educate the children of poor agricultural workers too, but those kids still speak Japanese and Korean at home! If the goal is to measure school quality, is it really fair to just put those kids into the average and fault American schools for doing worse than Japanese or Korean schools?
> I honestly think that the US "unique circumstances" are cope for the lack of decent income mobility and social safety nets that prevent a larger proportion of people from realising their potential.
Even if that were true, that would be more a point about the fairness of U.S. society rather than the quality of the educational system. I don't think it makes sense to conflate those two questions in a discussion of the U.S.'s competitiveness against China.
Moreover, income mobility in the U.S. doesn't break down by sub-population the way you might think. For example, while Hispanics have lower incomes because most are immigrants or children of immigrants, they have higher income mobility: https://economics.princeton.edu/working-papers/intergenerati.... Children of Guatemalan immigrants in the U.S. have higher income mobility than children of native-born Americans. Household incomes for Hispanics converges on the household income for whites within a few generations: https://academic.oup.com/qje/article/135/2/711/5687353
So focusing on PISA scores for "whites" isn't really about race or culture. It's just a proxy for "people whose families have been in the U.S. long enough to dispel the effect of immigration filters." If you were conducting the same analysis 100 years ago, you might try to exclude Italians or Irish from the analysis. Again, the point is to compare schools, not all the other sociological factors that are involved when dealing with immigrant populations.