Really? Let's take a closer look. Of course, if you look at the breakdown of the scores, which is available at the THES website for subscribers, you will realise that NUS does very well in criteria like 'international faculty', 'international students' and 'peer review' and rather poorly on things like 'citations per faculty'.
In terms of international faculty and international students, NUS and NTU have some of the highest scores. However, the high marks given are misleading. Allow me to explain why. In most universities, the presence of international students is regarded as a good thing because they correlate to the quality of the university. The general idea is that the better the university is, more foreign students would be willing to pay the extra fees to go there. For example, a Singaporean students looking to study in the US would generally prefer to go to the best place he or she qualifies for given the very high cost of education. People naturally prefer some place like Berkeley or MIT over Podunk University, if they can get in. The idea is that since you're paying so much for your education, you would want a place that gives you more bang for your bucks. Therefore, it is quite natural for THES to accord a significant positive weight to the number of foreign students in a university since most universities do not go out of their way to bring in foreign students just for the sake of bringing them in.
Of course, the situation is not like that in Singapore. NUS and NTU have a large number of foreign students not because of any intrinsic superior quality of their education but because the government effectively bribes a large number of foreign students with subsidised tertiary education and scholarship to study there. So, you see, the legions of foreign students in NUS are not a testimony to how great the university is but are really there because people are financially incentivised to go there. If you remove that component of that score, I suspect that NUS and NTU's rankings will drop drastically because few people would go to NUS paying full fees. That is not quite so for other universities such as LSE or UNSW with very high scores for international students. Foreign students go there paying full fees. Hence, their presence is a reliable indicator of the quality of these universities.
If you look at the 'citations per faculty' score, NUS does badly, which indicates that the quality of their research is not commensurate with their ranking. Of course, this ranking is not merely about the research performance of universities. Also, the 'citations per faculty' has a time lag factor and is calculated by looking at citation records from ESI over a 5-year period. To be fair to NUS, if there was a quantum leap in the quality of its research in the last two or three years, the ranking would not reflect that although it would show up in the ranking in 2009, provided that the same ranking criteria were used.