More than once in the last several years as a full-time radio technician, the question of accuracy of measurement of modulation came up and getting the right answer was pretty important for a particular application we were working on.
I ended up generating reference modulation signals using the Bessel Null function, employing precision audio and RF signal generators, (with the PC being the precision audio generator, with appropriate software), and then adjusting for the appropriate Bessel Null indications. Then I had a true "reference grade" modulated signal to use and see how the other equipment on the bench interpreted that signal.
It was both fun and a royal pain at the same time. But it gave me a great insight as to how modulation actually works. (deviation, actually, is the proper term, as modulation applies to AM signals and deviation applies to FM and other constant level signals such as phase modulation and various forms of phase shift keying.)
Actually making high resolution spectrum measurements of the modulated signal (0.1 Hz resolution) at various input audio deviation levels and frequencies led to some fascinating insights in how it all works. I'm not going to say I fully comprehend the math involved, though.
I realize this is a bit above the intent of this topic, but the science of RF is absolutely fascinating. The very idea that you can feed a small amount of electricity into a simple cicruit composed of capacitors, inductors, resistors, and a transistor or two and then be able to use a complementary circuit to receive that signal from many miles away is still, to me, magic. Magic that can be explained scientifically, but darn it, it's still magical. Science IS magic. At least until you fully understand it.