I’m confused. Please explain.

OP,

I would consider @sudnf’s point again. He (we) are advocating a method to make sure your system is reproducing the input recording as closely as possible. 
 

If one chooses audio equipment by what sounds good at the moment you can end up on a wild goose chase. I know, I’ve been there. Wow, that flashy presentation of my test disks sounds great! But, as you choose components that way, your system can start sounding flashy and soulless… and / or makes one genre of music sound great at the expense of others, and / or highly detailed but without the sounding natural like music. So training your ear to the real thing can give you an unchanging standard to shoot for. It can’t be electronically reproduced music, since that adds a whole new set of arbitrary layers between the instrument, other electronics and speakers. So acoustic music is the way to tie your judgement to reality.

I typically bring up the symphony orchestra because it produces sounds from the very edge of perception to overloading the hearing… very high db… unamplified. It also has single instruments as well as massed and covers the audio spectrum. But during the twenty years I got serious about training my ears to what the real thing sounds like, I also listened to lots of pianos, unamplified jazz, etc. all these help you create internal ruler with which to judge audio systems. There are lots of simple mostly two microphone recordings of symphony orchestras are available that naturally capture the venue… and all the subtitles.

When you start out, you know how to identify few of the dozens of attributes of music… so, if you learn what the real thing sounds like, you are more likely to get a system that reproduces variables you don’t know about yet.

When I started to listen seriously to un-amplified music to learn what it sounded like, my audio choices got much better. All genera of music improved at once, and the attributes I couldn’t put my finger on like rhythm and pace got better… it got more involving.