All joking aside, the simple fact is that if an amplifier produces different outputs when driven by different radios, then the radios are not driving the amplifier in the same way. The signal from the two radios are not the same and produce different amplifier outputs, power levels, percentages of modulation, possible spurious and harmonic components.
Harmonic components are related to frequency, multiples of a frequency. Spurious components are unwanted/undesired variations not related to frequency, just some 'freaky' product of a transmitter. Neither are related to modulation although a modulated signal, having more frequency variations than an unmodulated one, can certainly produce unwanted harmonics. The use of a sine wave as an input gives you the ability to 'see' any harmonics because of it's predictable 'shape' if you are using a scope to test with (a super good 'ear' can/could also distinguish a difference in it's sound, sometimes, but not as accurately).
It also helps if the same 'unit' of measurement is used. Changing between units of measurement introduces another possible variation in accuracy from the inaccuracy of that 'unit' conversion. Huh? Which means if you start with 'average' or 'RMS' watts, stick with that till you get through. Don't try converting to 'Pep' in the middle, see what I mean? Just an added hassle in remembering which 'unit' you are dealing with, and where/when.
Also remember that you are dealing with things that are mass produced and have variations in their itty-bitty components/parts, even the same parts. So, there will be some variation in how they all work together, which also means that two exact same radios can have different outputs. Not a huge difference, but still some. One of them 'ballpark' thingys, and there are always different sized ball parks. If there's a huge difference in how far away that outfield fence is, then you know you're in the wrong ball park, sort of. Make sense?
- 'Doc
Then there's always that mysterious 'thing' that always crops up in most testing, the difference you just can't explain away. Usually nothing really exotic, just means you haven't thought of what the problem can be... yet. That's the real head thumper! "Why..'thump'..didn't I..'thump'..think of..'thump'..that!?..'thump' ", ('thump' = head on table)...