H-m-m-m. Let us first get the horse in front of the cart. The definitive comment on loudspeaker performance is how it sounds - that is a subjective judgement. Ideally, one would want the loudspeaker to be a "transparent", neutral reproducer, so we get to hear what the artists created for us. If we don't like what we hear, then we can play with tone controls, equalization, to see if it can be improved. That is the basis of the audio industry.
The problem with subjective judgments is that humans are influenced by much more than the sound. Appearance, price, brand, other opinions and reviews all exercise bias in forming our opinions - the evidence is abundant. The worst subjective evaluations are those conducted in the fully sighted,"take it home and listen to it" method used by most reviewers and virtually all a casual listeners. There are no references, and adaptation sets it. Some call it "breaking in", but it is humans adapting, not electronic or mechanical devices changing. This practice has been the basis of audio opinions from day one, and it has led to the notion that we all must find the loudspeaker we like, because we all "hear differently".
Proof that this is rubbish was revealed to me in 1966 when I conducted my first blind, loudness equalized, four-loudspeaker comparison test at the National Research Council of Canada. The loudspeakers of that era were very distinctively "colored", not at all neutral, but most of the listeners in the group agreed on what was good sounding, even those who criticized the loudspeakers they selected and lived with at home. When they heard something "better" they preferred it. Where were the personal preferences? The second important observation was that the loudspeakers most preferred had the best looking - i.e. flattest and smoothest - anechoic frequency responses. These logically should be the most neutral. All electronics, even then, had ruler flat frequency responses. Why would loudspeakers be different?
There began a research career that extended over 5 decades, many papers, three books, an industry loudspeaker measurement standard, and many fundamentally neutral loudspeakers in the marketplace. For those willing to read, the scientific proofs are there, and have been for many years. An hour and 14 minutes will give a good summary of some of the key science:
Now we can recognize a neutral loudspeaker by inspecting the right set of anechoic measurements. This does not ensure satisfaction because recordings vary - they are created in recording control and mastering rooms by people listening through unknown loudspeakers in unknown rooms. It is wrong to assume that recordings are flawless, and all audible faults are attributable to the playback apparatus. Stereo soundstage and imaging are primarily determined in the recording control room, and there are no standards. Playback apparatus and rooms matter, but at a secondary level.
Except: A fact that cannot be ignored is that bass accounts for about 30% of one's overall evaluation of sound quality, and bass sound quality is dominated by the listening room - and they are all different. Arguably this is the weakest link in sound reproduction, but it is often ignored. There are excellent multiple subwoofer solutions to room resonance problems, but most people think that a bigger sub is the solution - or bigger tower speakers. Wrong.
Last several posts the use of single speakers for the majority of speaker tests. 
