Let me offer a thought: Human information processing is highly limited at about 60 bit/sec.
Hence, we experience our environment vectorially, i.e., we focus on variation.
Variation is not defined as every absolute change, but every change of pattern. Certain patterns are intuitive (eg, variation of leaf shape within a species) and within a continuum, others are not (e.g, when a caterpillar eats into a leaf) - a ‘combinatorial’ boundary is broken.
In analogy, not all distortion is equal. Highly reproducible distortion (narrow spec component) consistent with the fractale boundaries of sound generation seems very tolerable (and - we get used to it; we learn the pattern) yet when the boundary is broken (variation “outside” of the equation which defines boundaries of variation), we immediately notice and are - possibly - attracted, because our attention has been caught, or - often - disturbed, because the variation is random and ‘out of line’.
Short: variation does not equal variation. Distortion does not equal distortion. An argument based on specs alone is not helpful.
But then we all know that. We all hear differences in sound signature and presentation of different components.
Anybody more interested, I encourage to read up on ‘chaos’ and ‘fractale’ - the principles of phenotypic variation in nature. Applies very much to wave mechanics.