I have to admit I don’t know the science to this
As far as I know the -18 dbVu thing is related to analogue gear standard. After that point, circuit and components inside the gear would start to get warm and/or being overwhelmed by the electrical current, producing saturation, clipping and other kinds of distortion.
Thing is a lot of engineers started to perceive the distorted sound sometimes was a better pick for the piece, and then that is where the mojo and legend of analogue consoles started. That is why plugins that properly emulate analogue gear in general have a point where clean operation starts to distort (around -18dbVU, although the plugin’s manufacturer can change this), hopefully in pleasing ways.
That is also why a lot of oldschool engineers are so tight about gain staging (besides the sound to noise ratio): it is different to have all your channel faders up high producing distortion and the master channel fader down, than having your channel faders down and the master fader up. If you screwed the gain, you would have to manually adjust all the faders and probably lose the mix balance.
On the other hand it is the same thing in Ableton, for example, to have all your channels down by 6db and the master up by 6db. Digital gain does not provide any distortion, unless of course you hit the digital clipping point at 0db.
Engineers would also figure that specific channels on the console distorted differently (that is where a technology like Brainworx TMT finds its purpose).
also…sorry for the long long post!
tldr: working on music that is not yours is a great way to learn compression.
About compression, I must say I feel you! In the beginning I couldn’t compress anything “”“properly”“” in my music.
Thing about compression is that for me I could hear it better only on music from other people!
Maybe that is the case for @Gregg too, I don’t know, but please share! To me it was only after I’ve done some mastering for colleagues here in the Brazilian scene, first just for fun and then as someone who could actually improve the overall sound and balance, that I could start to hear and use compression properly in my own stuff.
In my own sound, I could never make my mind about it, and that was because it could never be better: only different. The composition mind set was never normative. If you’re discovering and searching for new sounds all the time, it becomes very hard to judge. It is all you.
When you receive a new piece of music from other people, it is way easier to judge! You bring in your experiences to contrast with the composer experiences. That gives you reference, makes you feel fulfilled when sound matches expectations, and make you feel uneasy where it doesn’t.
But now you can’t change almost any composition choices, and you start to value (a lot) EQ and compression.
You bring the snare fundamental down with EQ, because it was mudding the mix, clashing with the synth bass line. But now the power of the snare vanished. You bring a compressor and make so the it will avoid the kick and bass, but act on the snare. Tweak the attack and release, and now the compressor ducks everything by 2 or 3db after the snare hits. Now the bass and snare live without mud, but the pump after the snare will make it feel powerful again. Since you only ever brought things down, you now have a clearer sound and more headroom to make it louder. Win!
It is easier to use compression with instruments, because in the traditional instrument form, choice and function are more clearly connected, in general. By that I mean, you would almost never want a super resonant kick, for example, you would have gone for a tom, or maybe a tabla, or maybe a surdo. Texture, intensity, time and pitch are already determined by the composer/performer and by the instrument choice. So the engineer picks a alt rock, a punk or a jazz track and is more secure in making decisions.
Things become less normative on “avant garde” music where there is no defined “instrument” sounds. Then compression really must serve, even more, a creative outlet of the engineer. It is also the pursue of a feeling already in the music, as before, but the idea becomes harder to translate and accomplish. For a man that is a genius in that field, please check Rashad Becker.
After you develop your “mixing/mastering/engineer” persona with music from other people, you learn to switch with less problems from composer mode to engineer mode, and that is when you come back to your music and cringe hahaha! But cringe is good, cringe is great, because now you have found contrast and can start to bring in the right tools for the job 
