So suppose we have a mode of governance that is in opposition to western principles, let's call it f(n). Then through the perspective of your critics, they are focused on the worse case analysis, and going on about how the big-O of f(n) is bad. You can think of this as them making the case that if you take a mode of governance that violates western principles to its logical conclusion, it can only be <insert reason why it's bad>. I do not think they are wrong in that regard.
However, having a mode of governance that is in opposition to western principles doesn't mean that the worse case would (always) materialize. Think of it like quicksort, which in its worse case, is O(n^2). In practice however, we're more likely to get the average case, which is theta(nlogn). In this regard, you (and I) are looking at the average case of f(n), which happens to match the governance of Singapore more accurately.
Ultimately, anyone who has learnt about asymptotic analysis should know that it's a theoretical framework that aims to provide a qualitative assessment of the runtime of an algorithm. While it's useful, there are various factors in real life that could contribute to an end result that contradicts that assessment (processor speed, parallelism, being cache-oblivious etc). The dispute between the critics and their opponents is analogous to that, with critics who have never been to Singapore viewing the country as an authoritarian state (theoretical worse case analysis), while those who live there would struggle to reconcile their views (average case analysis) with that of their critics.
I hope this semi-shitty analogy does put things in perspective for various readers who are trying to make sense of the opposition between these two camps.