I think 5% sounding bad is a symptom that all recordings are being reproduced sub optimally. At least in my experience, as I shifted the overall balance such that the emphasis of different parts of the audio spectrum came in line, then all recordings sounded better and the outliers did not sound shrill or annoying. You could still tell they were bad recordings, but they were not offensive.
Pursuing the high end has an nearly endless learning curve. Thirty years in I was nearly clueless after thousands of hours of learning. Now at fifty years… I hear levels of nuance I was completely clueless of twenty years ago. For instance… this subject… I completely understand… it is a subtle but really important parameter in realistic sound.

