The problem is the difficulty in testing intelligence, and even what we mean by the term. Consequently, any such testing or investigating done, will inevitably be fundamentally flawed. However, once we label a group more or less intelligent, and rubber-stamp it as 'scientific', we run the risk of Untermenschen and Ubermensch. Popularly what is scientific is taken as what is valid or true, so the result of such research would do more harm than good - either fuelling othering and hatred, or inferiority complexes, or systematise unequal treatment in the name of privilege or correction thereof.
Quite simply, most of our tests are designed for Western European civilisation. If we apply it to others, we translate it or so, but the cultural biases are not similarly removed. For instance, if a culture uses inclusive counting or categorises things in a different manner, many verbal intelligence questions framed by Westerners will fail.
Similarly languages whose numbers correspond to their mathematical systems have been shown to greatly facilitate mathematical reasoning - thus biasing any intelligence test, in that any mathematical problem they have undue advantage. This may account for the good Asian showing, but is an artifact of the testing process itself. In short, it is cultural or even linguistic, but as those correspond to 'racial' groups, how are we to separate nature from nurture?
The question is what is Intelligence? Is it merely problem solving, that is dependant on the problems asked to solve? Or is it an ineffable quality? Testing for Intelligence presupposes too much, and thus would do societal harm without established benefit that we are testing for a real useful thing. By all means tests can be devised, then scored and checked if the pragmatic effect corresponds to any desirable outcomes - but to label this a protean term like 'intelligence' is not only invalid, it is foolish.
