HMMM, this is going to be a good thread, am curious what other folks are seeing too

Motors: Temperatures checked using one of those thermal sensors that look like  a thumb drive (accuracy?) This is basically only measureing case temperature after the flight as the heat is dispersing, core temepratures and operating temperatures likely are different than the temps shown here:
   BEST - 89F (ambient plus 2 deg)  (profile with a lightly loaded motor)
   More typical: 100F-110F (ambient plus 25-35 deg)
   WORST: I measured someone elses motor at 148F, I'm told magnet glue is only good for about 155-160F!

Battery: also measured using the thermal sensor - same discalimers  Inner cells likely a bit hotter, outer cells a little cooler.
  Typical: 110-120F
  Worst: I think Arch3's EF-1 Pylon Racer has gone as high as 150 (pulls 70A for 1:20 or so, NOT your typical Pro-Stunt run


ESC: Only data observed was using internal sensors built into ICE50.  Again, no idea how accurate, but probably pretty good plus consistant/repeatable:
   Best: I have had a hard time getting below 120F even on a profile.  Lower temps are going to require extreme measures...
   Typical: around 130-150F;  cycles up & down with the governing activity.
   Worst: I have heard of folks hitting 200F, but they quickly fix something so it does not happen again!

Well, read heat is around 700 - 900 degrees C, and I would certainly expect that in the case of a severe short circuit.  Once the battery bursts into flames you should be up to yellow or white heat, which is over 1000C.

Dennis:

If you really want to know about the accuracy of your thermal sensor, start your Google search with the phrase "thermal radiometry", or "thermography".  I used to work for FLIR, which makes thermal imagers.  There's a whole slew of things that can confound a measurement, the most prominent of which is a thing called "emissivity", more commonly known as "blackness".  Basically, the less emissive something is, the more it reflects its surroundings and the less it emits on its own -- so if you point that thermal imager at some polished aluminum it won't matter (much) how hot or cold it is -- you'll just be measuring the temperature of whatever your sensor is seeing in that particular mirror.

You can calibrate your sensor (you didn't realize you were firing off a Nerd Lecture, did you?) with a pot of boiling water and a glass of ice water.  Water is pretty darned black to most IR wavelengths, so the boiling water should show pretty close to 100C, and the ice water should show pretty close to 0C.

There's a whole science involved in knowing the emissivity of the surface whose temperature you're trying to measure and correcting the measurement for ambient temperature.  Good thermographers with strong mathematical backgrounds have tables and spreadsheets and probably these days little iPhone apps that let them do the correction.  Better (or at least more pragmatic) thermographers carry around little jars of flat black paint.  (Flat black is very emissive.  Very emissive surfaces don't need to be corrected for.  Guess what happens to the paint).

Edit:

I was wiki-walking last night, and was reminded of the difference between a theoretician and a practitioner.

Even smarter, and more pragmatic, thermographers use regular old black electrical tape (emissivity = 0.97, according to Wikipedia).  It's cheaper.  And easier.  And -- well, you get the point.

Be careful of shiny black -- it's only emissive if you're pointing straight at it -- otherwise you'll get cool glints off of your hot black surface.

Thanks, gents.  That gives me the temperatures to expect and how not to measure them. 

Sorry, Andy.  I shouldn't ask for help, then make fun of the answers I get, but that was too good a straight line to pass up.

Timm, it could be more in depth than simply cheap glue,,
The top of the line motors also have thermal limits in a similar range from what I have seen.
I think it most likely has to do with the thermal expansion rates of the aluminum can and magnets being different,,
( theory only, I have no substatiated documentation,, )

Although my question was about temperature ranges so I can deal with cooling and temperature measurement, I'm also curious about relative efficiencies of motor controllers and of motors.  If I can minimize battery energy going to heat the motor and controller, I can maybe get away with a smaller battery. 

So, are you going to buy a bazillion different motors and ESCs to test to see which ones are best?

No, I'm going to ask people who know.  That's how I design.

...so if you start gloating about all the battery weight you're saving before you weigh your ESC and motor, you may be disappointed in the end.

The full-service engineer is not only versed in the first law of thermodynamics, but can also add three numbers.