Interesting anecdote, but there had to be some other factor at work. For example, above a certain high coolant temperature, some emission control configuratations will restore vacuum advance at idle and low speed/load and/or suspend EGR, both of which will increase NOx. That's just an example. Since I'm not intimately familiar with the emission control equipment or configuration of that engine, I could only hazard a guess, but engines of that era had all kinds of klugy emission control hardware that can be a nightmare to sort out.

As a general rule for emission testing you want the engine as hot as possible. A change in operating temperature within the normal range - say 80 to 110 C will not cause that dramatic a change in NOx unless the high temp disabled the NOx control features; 2100 ppm would be representative of a non-emission controlled engine.

The high temperatures were effecting the NOx part of the emission control system in some way, but without a detailed understanding off all the emission control hardware and control systems it would be tough to sort out. Catalysts of that era were oxidizing only. NOx was controlled either with EGR, excess designed in valve overlap, or the spark timing map or a combination up to all three on on big, heavy cars. Excess valve overlap creates a full time EGR system, and GM used this strategy in the seventies, but it really killed low end torque and in town fuel economy.

Duke