I haven't gotten around to doing anything with the 10, but the genuine one has been hooked up to an RPi for a while, which is controlling a space heater.
That was still using a solderless breadboard, so it was an easy matter to swap in the 10 one by one and check if they were genuine. As was probably to be expected, they are all counterfeit.
They all seemed to be fairly consistent with each other and with the genuine one, although it turns out that these things are really sensitive to body heat--just holding one pinched between two fingers while slightly spreading the leads to fit the breadboard would heat it up 2-3 C. This made comparing different ones a bit confusing.
The genuine one seems to cool back down to room temperature noticeably faster than the counterfeits. I wonder if the genuine ones take more care to ensure that the die is not too insulated from the outside world so it will be more responsive?
Anyway, since I'm still using a solderless breadboard, and then things are designed to chain, it was not hard to rig it up so all 11 are hooked up at once [1]. (And yes, the resister is hooked up correctly. It is just a really bad angle in the photo that makes it look like it is off by one).
I've got a program running now that checks them all periodically and logs all the readings. Here are results after it has been running about 20 minutes:
22.437 [22.375, 22.25, 22.187, 22.312, 22.375, 22.187, 22.25, 22.25, 22.375, 22.25]