I have an even wilder speculation. I think they maintain a quota for international students, ie. two koreans, two chinese, a bunch of new zealanders, etc. (I made up the numbers). So we are not actually competing with the vast majority of domestic applicants, and we are evaluated based on a very small international applicants pool, the relative strength of which determines our chance. My empirical observations seem to back this theory up. But again, that's my pure speculation. Wish us all good luck!
fellow international JS1s: during my interview I asked JS about when should I expect a decision, and she told me that since there are a lot of files it might take some time. At first I thought it was because I was not good enough (which might still be the case), but now I think such timing might apply to all international JS1s in this batch. She might want to go through the whole held application pool (not just internationals) to decide how many internationals she would take before issuing any international JS2s.
My theory is that they want to estimate the LSAT median for the incoming class, and then sprinkle internationals into the mix until the median = 173 again. We're the easiest way to remedy a low LSAT median, since there are a lot of us at 173 or above, and we don't mess with the GPA quartiles.
I am fairly sure you are right that we are competing against other internationals (or even with our own people). But it does not explain why no international from the latest JS1s gets a JS2 (or maybe someone did but not on TLS, but I highly doubt that): I am sure JS has finished reading the materials from true international before issuing the latest JS1 invitations (just like she likely finished reading all URM application before the latest batch of JS1).