He's kind of mixing apples and oranges in his article. On the analog side, the loudness to softest has been brought together by compression. In digital, the inner soft music has been almost eliminated, leaving the louder, outer core of the music, then it's what you do with the outer core.
If we think of going to a rock concert and I drop my car keys, would you hear it? It made a noise, but the level of loudness of the music playing totally masked it. This is a natural analog "hide" if you will (as the author called it). How much of the nuance is actually in a typical recording (that's not highly compressed) is for debate, but I'm betting that there is lots of it and it may be some of the things that gives a recording it's "sparkle", especially if it sneaks thru some of the quieter passages.
Then there is the problem of the bands final mix into stereo from all the recording sessions. They like how it sounds, but the record label management decides to release it in 3 formats: CD, MP3 and maybe vinyl, each with their own set of "mechanical" problems of getting it onto the media. The vinyl cutter may decide that the music is way too hot for the vinyl so he has to do some compression to keep the stylus from jumping out of the grooves. The MP3 version had all of it's underbelly scraped away, and the CD was made to blast, "normalizing" it, even red lining it.
Main stream music has made another product, until the audiophiles listen to it. Failure by our standards, success by the record company's standards.
Wayner