So as I understand it, you are not taking the "extreme" view that sighted impressions are worthless, at least for the individual.
But you edge towards the idea it is "useless" to gain information about speakers from the subjective impressions reported by other people (e.g. audio reviewers, presumably other audiophiles...)
Useless in the context of reviews of audio components for consumers. Indeed, the fact that the bass is sticking out in one reviewers system/room with a given component is unlikely to tell you something useful for another user. Again, this is because the sighted observation includes both a real signal and many fairly random and uninvestigated additions (room effects, biases, mood, hearing, etc.).
Imagine, reductio, that each subjective review was derived from a true signal plus a randomly selected frequency suppression/enhancement curve. Would that be useful? That's more or less what these sighted reviews are.
Of course I take your point, we all know speakers will sound different to one degree or another in different rooms, and to degrees, to different listeners. Nonetheless I think you are exaggerating that problem insofar as you imply it's so insurmountable as to be "useless."
Now you can quibble about mostly useless or totally useless (allowing the idea there *may* be some shared tastes).
I wouldn't call audiophiles swapping impressions of speakers (reviewers included) even "mostly useless." I know it can be quite useful - I've been guided by the reports of other audiophiles and reviewers to plenty of products that I loved. Of course you can approach using other people's impressions in a way that, yeah, they'll be useless. But you can also be more careful in collating impressions about gear. For instance, when I see someone describing the sound of a speaker I notice whether they are good at putting subjective impressions in to words, do their descriptions of speakers I've heard match my own impressions? Do they seem to care about and listen for the things that I care about? Etc. When there is a decent amount of "subjective data" available that seems to have lots of arrows pointing in the right direction - "that sounds just what I'm looking for!" - I've rarely been shocked or disappointed when I heard the gear in question.
And it goes both ways. I've lost count of the number of audiophiles who told me my descriptions of speakers matched their own impressions "that's just what I heard!" In fact just recently, again, on another forum a member who owned Harbeth Super HL5plus speakers had the opportunity to buy used Joseph Audio Perspective speakers and was asking if anyone had heard both. As it happened, I owned both speakers! So I wrote a detailed description comparing the sound of each. That member bought the Joseph speakers and reported afterwards that everything I wrote about their sonic differences/characteristics was "spot on." His own lengthy description comparing the two speakers once he had the Josephs matched my own experience with them perfectly. Similarly, someone else just bought the Joseph Speakers based on my writing about them, and was thrilled - sounded just as I'd described them.
And one reason, I submit, that careful sonic descriptions/impressions can be useful is they can convey what aren't always obvious from the measurements.
Certainly any audible characteristic WILL be in the measurements, but given all the different picky sonic characteristics many audiophiles are interested in, unless an individual has a good technical grasp of speaker measurements, if not LOTS of personal experience correlating all the different variations in speaker measurements to the subjective character, then the short hand sonic descriptions can actually do the job "better" in many such cases. The measurements don't have something written in them "And It Sounds Like This!" That's where, in so many cases in life, intersubjective information comes in handy - another subjective listener can explain "It Sounds Like THIS."
And a problem is that when you DO have people who have the technical chops to read and correlate measurements to predict the sound, that person doesn't necessarily share the taste or personal criteria of another audiophile reader, and may gloss over characteristics that the reader would care more about. Or the technically inclined person may not have great facility translating the measurements into subjective correlates, or in to language, or may simply be less interested in doing so. In other words: The very engineer-oriented types who are in position to understand the measurements are often the ones who view subjective descriptions with suspicion and so won't put much in to such descriptions of "How It Sounds"! So...you don't get much actual subjective detail. (That's one thing valuable about John Atkinson - he has the technical chops, and also seeks to correlate them with sonic descriptions of the subjective effects).
So for example, the Devore O/96 speakers that I really love. The Devore speakers have critiqued on ASR before, and in the comment section for the Stereophile Devore O/96 review some clearly technically knowledgeable folks just tore them apart for their purported deviations from "best practices" as they saw it for speaker design. And yet NONE of the technical information they used for their arguments gave me any hint that this could be a speaker I might actually like!
Rather, it was carefully reading a whole bunch of different reviews of those speakers, noting that there were some constant themes about the characteristic of the sound being described, that are just the type of characteristics I value. And upon hearing them...yup!...the descriptions were right. They were doing all the things described, and I loved the sound. (Again, this has been the case for numerous speakers I've really loved).
There are far more audiophiles who have made satisfying purchases based on swapping gear impressions or on reviews than your reply suggests would be the case. Even Steve Guttenberg, not that I'm a huge fan, has his youtube comments often filled with people thankful for his reviews, which led them to products they are thrilled with.
And that also speaks to the idea that people hearing speakers in different rooms becomes some insurmountable challenge for exchanging sonic descriptions.
For one thing, audiophiles (and reviewers) tend to experiment a lot with speaker placement to optimize the sound of whatever speaker they buy. Very often this will be part of the description - "found they sounded bloated too easily unless far out from the back wall" or "could be placed pretty close to the back wall without bass bloat" etc. Yes, there is a level of "noise" when talking about the same speaker heard in different rooms. But if a reviewer is careful, and especially if numerous reviews or audiophile reports tend to converge, that can be informative. Are most able to get a similar sound, or not? If they are, that tells you something too.
As I've often said, a tool is only useful if you use it. Many here dismiss the usefulness of subjective uncontrolled reports even for speakers, and that's fine for any individual. But it doesn't mean others can't find that form of information significantly useful.