Even Tim who has never seen a measurement that he didn't like has had to return products that he bought based on their measurements/specifications because they didn't sound good in his home.
Just a clarification, mep; I don't think I've ever ordered a piece based on measurements alone. I don't trust manufacturers' measurements and comprehensive, independent measurements are very hard to come by. Most of the purchases had more to do with reputation, design philosophy and reports from users. More often than not, I returned stuff not because it sounded bad, but because it simply didn't earn its keep. It had a price tag but made no audible improvement. When you balance the volume precisely between your existing, good quality component and compare it blind to the one you're auditioning, you'd be surprised how often you can't even hear a difference between them. At least that's the case when neutrality is your goal and the goal of the people designing the components. If a signature sound is the goal, all bets are off, but I don't play in that league.
But your larger point is correct. My ears are thoroughly engaged in the process.
The Japanese predicted that American audiophiles would make a purchasing decision based on how many zeros came after the decimal point in their distortion measurements and they were right. Why settle for .0001% THD when someone else is offering .00000001% distortion?
It was worse than that, Mep. Most of the time they were measuring that THD at one isolated point in the frequency spectrum and at a load that wouldn't have been a challenge to a strong headphone amp. Thus my distrust of manufacturers' measurements.
If we could deduce exactly how something was going to sound in our house by simply looking at measurements, we would have never had the need for dealers to demo equipment.
I wouldn't go quite that far. Something as complex and mechanical as a speaker or a turntable is going to be hard to predict, but given comprehensive, independent measurements verified by a second party, like the measurements of the Benchmark from Stereophile I linked to this morning, verified by the measurements from The Audio Critic at that same link, I think you can get very, very close. In fact, I think you can pretty confidently conclude from those measurements that the Benchmark is not going to
sound at all. And that is, of course, its job.
Oh, and by the way, I've met several measurements I didn't like -- wow, flutter, rumble, crosstalk, inner groove distortion -- I didn't like any of these measurements, so I eliminated the need for them altogether

.
Tim