agree in general, but let's remember also that the entire heated surface is not at the same temperature and that there are local hot spots which exceed the temperature sensed by the thermostat prior to its opening the circuit. Before I banned all "non-stick" coated cooking utensils in my realm, I had a Presto brand griddle and it was evident from looking at the surface exactly where the heating element was - the non-stick was discolored and thinner in the immediate vicinity of the location of the serpentine heating element. So, while the entire surface doesn't react, I believe there are local hot spots, in those regions closest to the heating element, where decomposition occurs. The thermostat is relatively remotely located, and the surfaces most often are aluminum, a less than optimal thermal conductor. For the pure fluorocarbons , I agree they're highly thermally stable. But we should not overlook the effects of repeated heat cyclings and the presence of aluminum as substrate which reacts readily with fluorine - that's the test - how stable is PTFE & analogs in contact with aluminum. Then, the next step is to determine any catalytic effect those breakdown products may have on surrounding PTFE coatings. Typically free-radical mechanisms are involved, leading to chain-propogation. That old griddle was nearly bare above the heating element versus the remainder of the surface.