I think you have a good point here.
Objectively good performance is easy to see. Nice smooth, straight lines all around is what we look for.
Objectively bad but subjectively good (or at least, sometimes euphonic) performance is harder to see, especially because directivity is often a mess in these situations.
So you have to mentally piece together where you'd want to hear boosts or cuts... net that out with resonances and PIR... and create some kind of mental image of what music might sound like put through that filter, and decide if it would be nice or not. That is extremely hard to do in a reliable way. Tuning an objectively bad speaker to sound good is possibly even easier done by ear than by measurement... probably why DeVore does it that way.
I feel like I could make a vague guess if you put a gun to my head, but finding diamonds in the rough by reading graphs is certainly "advanced topics" at best. Case in point, once in a while even Amir will listen to a speaker and report being pleasantly surprised after seeing some gnarly graphs.
Yeah, this brings in an issue I've mentioned before about the usefulness of speaker measurements.
Among the approaches are two somewhat different concerns regarding measurements:
1. Using measurement graphs as a way of deciding between the range of "good" and "bad" designs.
2. Using measurements to tell you
what any particular speaker will sound like.
#1 is generally the approach taken here. And it's useful. You take a standard, with good engineering and evidence behind it, about what type of speaker curve/measurements "sound good" and you have those as your template to evaluate speakers against. Then you weed out "good" from "bad" designs against that template. And for the newbie it's easier too, because you can point to, say, the Harman curve and say "look for measurements like this, it will sound neutral and smooth in most rooms." You don't even have to know personally what the curve sounds like, you can just use it to end up with a speaker with those measurements...that sounds good.
But it's an entirely different ball game to move from "a speaker with this curve will sound good" to actually knowing what any particular speaker will sound like just by looking at the measurements. Because as we know speaker designs available to the public are all over the map, and can sound quite different in a myriad number of ways, for all sorts of reasons, on axis response, off axis response, resonance, and often various anomalies combined. THEN we are talking about having lots of personal experience correlating various precise measurements to their sonic consequences to predict what a speaker will sound like. And even people with tons of this experience still tell us it mostly gets one in the ballpark, and there can be some surprises in comparing listening impressions to some measurements.
So, all is fine and good IF the audiophile only cares about speakers that fit that certain paradigm declared as "measuring well." But IF you are an audiophile who really likes all type of speakers, and enjoys different presentations, that approach isn't enough. You either have to get REALLY good and experienced correlating measurements to speaker performance, or...you gotta go listen for yourself. And there can be a problem relying on what some more technically experienced people will say, because very often such folks have strong opinions about what is good or bad, and they will dismiss a speaker as "don't bother with this one" when it may well be a speaker you would really have enjoyed.
Personally, I can look at the measurements of a speaker like the Devore O/96 and get a general gist of some of the things I might hear. But I can not perfectly predict how all the frequency response/off axis behaviour/resonances will combine in the total sound in a specific room. Will that resonance bother me? it might. But it might not.
Upon hearing the O/96s, I got my answer: I found that all together, measurable idiosyncrasies and all, the sound was really compelling...transfixing even, for my taste.
There's a similar divide in the approach of the subjective rags like Stereophile vs ASR. Like the difference between 1 and 2 above. ASR will measure along a "good" or "bad" paradigm based on specific design/measurement goals. A mag like Stereophile attempts not so much to say "good" or "bad," but rather...there's a whole bunch of differing speaker designs out there, and THIS is how X speaker sounds, and THIS is how Y speaker sounds...and the reader can decide if they are the type of characteristics that would interest him/her. So, a wider...or another type...of audience. Yes, I know all the criticisms lobbed at subjective reviewers. But I'm glad there are both approaches available. The ASR approach works for many audiophiles, but so does the Stereophile approach for others, interested in speakers ASR would never approve of.
