Ok folks, here's what I came up with.
I ran the Kill a watt (referred to at this point as the kw) through a few tests to see how it would perform and survive.
I held it at 125 volts and put a 1000 watt resistive load at first on it. I used a dedicated calibrated power supply and as the wattage started climbing, adjusted the voltage accordingly to put 125 volts out TO the kw meter to see how it read it. Overall though the voltage drop was maybe a volt from 0 to 1000 watts,
the meter tracked pretty linear and accurately, being off by maybe a few tenths of a volt here and there. Nothing significant given the specs on the meter.
I then ran the watts down to 500 watts and ran the power factor from -.4 to +.4 to see how it tracked under inductive and capacitive loads. (Note: I didnt run it at the full 1000 watts because I was too damned lazy to go look for the cap bank and the other transformer to hook through to control and bring the power factor down that far. Again it tracked voltage and power pretty close.
We have a small heating oven which we use for bearings sometimes, and next step (without getting into all kinds of geekly how I set it up stuff) I placed the kw into the oven to start heating it a bit. I set it to 1000 watts of load through a resistive load which would be a unity 1.0 Power Factor and let it run. I ran it up in 5 degree increments holding for 5 minutes at each step to 180 degrees F. At about 165 degrees the meter started goofing and acting strange so I stopped the test, however up until the point where it started blacking and scrambling it was tracking pretty much dead on.
Conclusion, the meter will track accurately through temperature changes and load changes and the only way heat seems to affect it is when it gets wayyyy too hot, much hotter than you really should be running the thing, it will start freezing, scrambling or blacking out on the display.
Nother note, at the higher temps I used an IR thermometer and found that around the plug was getting MUCH hotter than the rest of the meter, perhaps the heat was causing the metal to flex a bit and a looser connection at the plug in, hence the extra heat..... also not checked BUT if that did happen, the heat could have easily radiated up the metal prongs of the plug, into the meter workings and caused much hotter temps than the 165 ambient I was subjecting it to and giving it problems there too.
Again though the thing worked as it should until I just ... stopped. In other words there was no 'drift' in the readings or gross inaccuracies as it went on.
For what this was worth, it was fun to do at least and I hope this experiment helps somehow.
Aaron