A few small data points
Coils generate their own heat. Likely not as much as the combustion in the engine does, but then the outside of the engine is not as hot as the insides of the cylinders. I mistakenly left an ignition switched on for an hour-ish (engine not running) and the coil was too hot to touch. But it still worked.
Bentley mounted their Lucas coil on top of the radiator. I've overheated that car pretty well in Texas heat, with no signs of coil damage.
I had a Bosch coil fail on the racecar, which was fun (luckily it was the last lap of a practice session). It acted exactly like the car was running out of gas: it would accelerate happily but as it approached redline it would suddenly slow as if the throttle had closed. Shift to the next gear and it would accelerate again for a bit, then slow. I believe that the half-second that it took to shift gears allowed the coil to cool enough to give a few more seconds of sparks. FWIW, the coil was mounted on the firewall in that car. I do not know if a coil can recover from that. I wasn't willing to take a chance.
I recall seeing a major coil overheating incident during a NASCAR race once -- the top of the coil (located in the cockpit) blew off, spewing the oil filling. I don't recall what type of track it was, but NASCAR "generally" runs more or less full-throttle, full-time, in one gear. Running below redline generates less heat than running at full throttle.