Amplifier designers do have the proper equipment as to get an FCC certification (required in the U.S. for HF amplifiers) they perform their own testing and provide detailed lab reports to the FCC. You would not normally use an "attenuator" for this, as that would be a really BIG attenuator. Usually you use a directional coupler with -20 dB (or more) decoupled port outputs and then use smaller attenuators after that to get down to the +10 dBm (10 mW) level or so for the spectrum analyzer. Way cheaper, and less hardware (and a lot less heat in the attenuator). I can make these measurements here on my bench and have all the hardware; but it's impossible to keep all that in the hamshack and active and also a bit fruitless. Takes two very low distortion signal sources, directional couplers, attenuators, very linear mixer, and a spectrum analyzer you can trust to be accurately calibrated and operating at low enough signal levels to not introduce its own distortion. Takes me a while to even set up to do this. In my experience, most SS amplifiers aren't great. GG power triode tube amps can be really good if properly loaded and operated below compression. My own 4-1000A grid-driven tetrode amp is "fair," but not as good as, say, an 8877 triode amp at the same power level. The "12Vdc" operated SS amps are about the worst, and unfortunately that's what's in most of our rigs!