What is purity?

We've been conditioned to believe distortion products in audio equipment are important. Else, why would manufacturers and reviewers go to such great lengths to measure and write about them? I won't bore you with the history behind the story: audio pioneers discovering what mattered, manufacturers stretching discoveries and measurements to differentiate their products, and the myths and truths that evolved. But let me suggest the truth is a matter of degree. Of course we can hear gross distortion, and even tell the difference between music 99% pure vs. 99.9%, but beyond that? Does it seem reasonable that differences of 99.99% (0.01% THD) and 99.95% purity (0.015% THD) are easily identifiable in home audio systems? These differences are tiny indeed, and the distortion products are at an almost inaudible level, yet most believe they are not only relevant, but critically important. Sometimes it's good to put things in perspective with a laugh. I remember comedian George Carlin making fun of miniscule amounts of carcinogens in drinking water, asking his audience if they would drink if it were only 99% poisoned: I don't care, I drink it anyway. You know why? 'Cause I'm an American and I expect a little cancer in my food and water. I'm a loyal American and I'm not happy unless I let government and industry poison me a little bit every day. Solid state devotees strive for falling distortion levels - the lower the better - because it is believed odd order distortion is to be avoided at all costs. And tube lovers believe the opposite: a little second harmonic distortion is beneficial in their daily diet of music - and higher levels are actually coveted. At least that's the myth. But is this distortion responsible for the "tube" or "solid state" sound? I struggle with that belief. Modern preamps, both tube and solid state, have low distortion, typically under 0.1% (99.9% pure). Are we being asked to believe the addition or subtraction of 1/10th of a percent of overtones of one type are so different the sound changes dramatically? I am convinced the differences between solid state and tubes, as pronounced as they are, are due not to one characteristic but to many. That may seem obvious or overly simplistic, but there's much we can dig in and learn.