At what point do the measurement numbers matter? Are they valuable as an explanation tool when something sounds great? Or the opposite?
If I use measurements as a history lesson, where I look at them only after listening, I might get one outcome. For example, a new DAC. Upon listening its audible performance will fit into one of several categories: good, meh, bad.
Let's imagine that after listening I discover the measurements show the DAC to have extraordinarily low noise and distortion. If the sonic results are good, then I can say with some confidence those low noise and distortion measurements surely contributed to that result. If it sounded meh, then I can place it in a so what category. And, if it sounded dreadful, then I simply suggest there must be another measurement that could have explained it.
On the other side of the coin, if I use measurements as a predictor of the future, then I begin my listening journey with a set of expectations. That same DAC, with its low distortion and noise, should sound great! And, if it were to meet my expectations then my predictive capabilities—based on those measurements—are confirmed. If it doesn't match, I don't dismiss my predictive prowess, I just simply suggest to myself something's missing.
Any way you look at it, the numbers matter but only if we narrow the focus down to something very specific.
Broad returns on narrow measurements don't tell us much.
