You’re oversimplifying things, and coming to the wrong conclusion. The literature from automated software makers, such as Audyssey, makes it seem like it they can make any speaker sound wonderful at the push of a button. It’s the audio equivalent of making a silk purse out of a sow’s ear – making Madonna sound like Lady Gaga. No software can do that.
Although there are plenty of good sounding speakers, there are probably more that generate distortion or harsh sounding resonances – essentially unwanted noise. They are poorly designed or built, and no amount of electronic equalization can correct that. The sources of noise must be eliminated or minimized by proper speaker design.
The goal is to make a speaker with a flat frequency response, low distortion, with little or no noise or resonance. This sound should also be as widely dispersed as possible. Flat frequency response should be heard both while directly in front of the speaker (on-axis) and while off at a fairly wide angle (off-axis). Widely dispersed sound is responsible for generating convincing sounding images of musicians. All of this results only from good speaker design. No software can correct for poor speaker design.
Assuming we start with good sounding loudspeakers, the sounds we hear are a complex combination of sound coming directly from the speaker and sounds that reflect off of floor, ceiling and walls. The behavior of these direct and reflected sound waves varies widely depending on the speaker’s and listener’s location in the room. The direct and reflected waves can add together if they are in-phase with each other, or cancel each other out when out-of-phase. This behavior also varies widely depending on a sound’s wavelength. At some wavelengths direct and reflected sounds can arrive in-phase at a listener, and at other wavelengths they can arrive out-of-phase. This isn’t so simple that a computer algorithm can sort it all out.
Most methods of electronic equalization that I've heard can improve on-axis sound, but at the expense of the off-axis sound. With speakers that have both on-axis and widely dispersed flat frequency response, I’ve noticed that Audyssey tends to “suck the life out of their sound”.
Automated software, such as Audyssey, can do a decent job with bass management and general set up a system of surround sound speakers directly from test signals and input from microphones. This shows some promise. But when it comes to equalizing a speaker’s room responses across the full audio spectrum without stomping on off-axis responses, those results have been mixed so far. With some rooms, some speakers, and some listening preferences room EQ can work to some people's satisfaction, but with other speakers, rooms, and listener preferences, it works poorly. I think of automated room EQ, at best, as a work in progress.