That comment doesn't pick up on how much of an impact the low quality of the data has on the outcomes. For instance the commentator points out that there should

That comment doesn't pick up on how much of an impact the low quality of the data has on the outcomes. For instance the commentator points out that there should be a larger difference between the average IQ's of Hong Kong and mainland China, given China's massive peasant population. However, the wikipedia page points out that the Chinese rural population was never tested, and that the differences in scores was artificially added by the authors: "For People's Republic of China, the authors used a figure of 109.4 for Shanghai and adjusted it down by an arbitrary 6 points because they believed the average across China's rural areas was probably less than that in Shanghai. Another figure from a study done in Beijing was not adjusted downwards. Those two studies formed the resultant score for China (PRC).