Well let's imagine 2 people in a room and a third person playing the piano. We're both hearing the same sound waves, or with very small and neglegible differences. That is our natural. When we hear the same sound file from the same headphones through the same sound chain, I assume whatever impressions we had would be the same, so if you thought it was distorted, or dark-sounding, I'd think the same. It doesn't mean we perceived the same sound, but the difference between both sounds (live piano vs listened through headphones) would be the same. Like if 2 people look at 2 pieces of paper painted in 2 shades of green, one brighter and the other one darker. We might seen green in different ways, how can we be sure we're actually 'seeing' the same color? However, I think we'd both agree on which one was the darkest and which one was brighter, since it's not about absolute sight, but differences between colors. It's a confusing example, but I think you can see my point.
Having completely neutral equipment is very expensive. The thing is, at a certain point you need to understand that paying 1000$ more on a better amp to get rid of that 1dB 'spike' just isn't worth it. Any equalizer will do the same. You can still get very good and very neutral gear with a small budget, for example you have the O2 amp which is very flat and measures amazingly, with everyone reporting how they simply don't hear the amp in the chain.
As for measurements, they're an important part of making decisions. You can have mixed opinions, while one says "sparkly and detailed treble", someone else says "bright and bass-less". A graph won't lie, it has no opinion. Some thing you can't get from measurements, like soundstage or some small aspects of sound, and on those we rely mostly on reviews, but I think a healthy combination of both objective and subjective input is the best way to go.