To me you have answered your own question here really. If standard, commonly accepted specifications and measurements were all we needed to design a good amplifier, why then it's simple. We set parameters on the test equipment and keep adjusting the numbers on the circuit till it hits the right thresholds or sits in the right band etc.
What you have missed out is, response time of a signal( at any given frequency), it's reaction to changes in a complex load (the speaker can behave quite differently at different frequencies, temperature, current and so on), impulse response, response to a massive need for current like when an orchestra all, suddenly, play at once and loudly and so on.
So what you are really saying is the two amplifiers seem to sound the same in two dimensions not taking into account that you live and hear in three.
THD+N, Watts at 1kHz blah blah. Might as well put your finger in the air really.
We usually do them right at the end for the user manual. Always seem to come out fine.
Haider (of
Sonneteer)