Pleasurably better, not measurably better

No?  First, I don't believe that "accuracy" is really possible in a speaker/room combination outside of the original recording space, nor do I believe the engineers would have mixed to reproduce that space.  Next, if you look at the research undertaken by JBL and Harman over time, speaker makers and measurements were based on human desire, not energy time curves.  Those who misunderstand that always end up with systems far too bright and dry.

Lastly why should I care about accuracy if it isn't more engaging to listen to?

I'm not buying a telescope to measure the distance or size or material composition of a distant star.  I'm not buying a tool to help me measure the amount of chlorine in a pool.  I'm buying gear that makes me feel good.  At best I can achieve a system that sounds good with a wide variety of source materials.

Couldn't agree more with this, not to mention all the circuitry that music goes through in the recording process should definitely change the sound of the musicians in the studio.

I cringe when the word “accuracy” is introduced. The conversation tends to head towards simplistic measurements and goes off the rails.

For me, I need the gestalt right… conversations about accuracy tend to go to a few single measurements that completely miss the point. I have been a scientist and technologist for all of my life and when coming to carefully reasoned and logical conclusion based on all sorts of detailed data, I always ask my team, “does this pass the laugh test?”… if you look at the question… and the conclusion… does it make you break out laughing? As in, that is ridiculous! If so, go back to the drawing board, because there is something wrong with your premises. Ultimately science is about observation, then postulating possible causes and relationships, and if your theory doesn’t fit reality… it is wrong.

Anyway. Bring up accuracy, and typically over-simplified models of the world seem to get proposed and the whole conversation seems to go off the rails. So, I try to stay away from the word. I feel my system captures the gestalt of a musical performance… but I suspect measuring some simple parameters might say otherwise.

Measurements are definitely helpful. But what everyone experiences as pleasant varies. And what we call pleasant we tend to define as good as well. There the trouble starts.

I´d never expect anyone to like my prefered sonic signature.

When I bought my first stereo the shop owner told me he had sold an amp/speaker combination that was horrible to his ears, but the buyer was a professional cellist and the setup reproduced his experience as a player sitting in the orchestra very well.

„Pleasurably better“ – that´s a good term, because it´s asking to define what exactly is pleasurable. I also like the „euphony“, literally translated as „good sound“. We know euphoria and that it´s experienced in individual ways, so here´s an audio specific sister word.

I think culture, background and training are at play here. Those deeply engrained in the scientific and engineering communities are disciplined to defend their positions with hard data. It’s part of the culture and for valid reasons, when the goal is to prevent airplanes from falling out of the sky, for example. Those among us who have developed highly sensitive "antennas" to various audio phenomena and emassed a farily extensive vocabulary to relate those observations to others are quite comfortable sans data. Attempts to minimize the attachment of data to subjective audio experiences often frustrates data-driven individuals. It’s understandable. You can’t blame a compass for pointing north.

There are no "good guys" or "bad guys" here. A little intellectual humility is a great way to begin (and, end) a conversation.

Good post, Erik. It’s been a pleasure reading the comments.

Those deeply engrained in the scientific and engineering communities are disciplined to defend their positions with hard data.

I think there are a lot of people who want to play at being scientists or engineers when they defend their technocratic positions on audio.  The giveaway is when they attempt to make claims about metrics that are simply not in evidence.

Lower THD for instance.  Not proven to sound better and under some circumstances higher may be preferred.  Like measuring the hardness of a metal.  Yes, we can measure hardness, but is harder better?  That's a different issue and judged by application.  Measuring hardness is an engineering discipline, but claiming it is universally better is outside the scope of the discipline.