Even in an anechoic chamber, there are poles and resonance effects in the speaker cabinet, the driver, and the crossover if one is used. You can't get to perfection that way. Besides, the absolute lack of reverberation in such a chamber is very disturbing to some people.
Define perfection, as applied to audio reproduction, anyway.
You might say that you want to hear the "original performance", well most recorded music is produced by laying down individual tracks and then mixing them, entirely electronically (or digitally). There was never a single "original performance " to reproduce.
Maybe, just maybe, if we talk about a live event ideally with musicians playing unamplified instruments (say, an orchestra) we could define it as reproducing the sound that you would get at some point in an auditorium and, with binaural recording and minimal pissing around with the signal chain you could achieve that. You could, at least, start to think about measuring the signal at source and again when reproduced to show that the reproduction is accurate.
As far as I know no review or comparison of audio equipment has ever set out to define what the reproduction
should sound like and compare reality with that "ideal".
After that it's all a question of my distortion sounds more pleasing to me than your distortion.
Edit: Actually that brings to mind a spoof article, possibly in Everyday Electronics in about 1979 or 1980, which lauded the new "octophonic" playback system which was capable of reproducing the exact sound experienced by some one in seat ZZ18 in the Albert hall (or wherever, I forget). It then went on to describe how each channel was devoted, in turn, to the sound of someone in the audience coughing, or rustling a bag of crisps etc until, finally, the 8th channel was devoted to the concert - but very quietly as seat ZZ18 was way at the back!