In professional audio, we use loudspeakers and the rooms they are in to make judgements about the quality of what was recorded and also to predict how a recording will sound when other people play it over other loudspeakers in other rooms. I think we can all agree that loudspeakers should do this accurately. But what does "accurate" in this context mean, anyway? Literal physical acoustical accuracy is out of the question - microphones and loudspeakers lose too much information for truly "accurate" reproduction to take place. So what we REALLY mean by accurate, it seems to me, is a playback that presents a compelling sensory illusion of the original event, an illusion that feels realistic, lifelike and, well, present in the same space with us. Also, we'd like the illusion to include the original space as well as the original sound source. When those things happen, we call the sound "accurate," or maybe "totally f___ing awesome!."

Meanwhile, we have to keep in mind the nature of the relationship between the loudspeaker and the recording/mixing engineer. As engineers, we all learn the sound of our favorite loudspeaker played in our room. As our craft develops, we learn how to anticipate what the implications of various sound qualities as heard by us will mean to others listening in other spaces, so that when we get our recording to sound "right" to us in our environment, we can predict that others will say "it sounds good" when they play it back in their environment. Then we can get paid.