Perhaps you should look at what I posted, not just that one sentence. I said sticking a thermocouple on the reactor, even poorly, would be good enough to see if the IR camera is reporting a value that is close to reality. And yes the thermocouple, type K, would melt at 1380 C.
Right. Besides, the heater wire would have burned through well below the reported max surface temp of the reactor. The researchers were, and still are, naive to the manipulations of AR. That doesn't necessarily mean they are bad scientists, but they did invalid work here and they should admit it or at least release the raw data at a minimum.