When I used to record and mix my own music it was the same way. Dynamic range compression isn't a bad thing if it's done right. I can tell you one thing though, the full dynamic range version played back over a car stereo with road and engine noise sounds worse. Our ears apply some of their own compression when louder sounds are involved anyways. The real problem with the loudness war is when its taken to an extreme, where the range is compressed to +-3dB and the level is cranked up to 0dBfs. Several albums mastered in this way have minor clipping issues. Often its less about the amount of compression and more the attack and release times used.
A good mix and master sounds good, even when compression is applied. It's not the be all end all.
I wonder, with all entertainment delivery formats being digital, why can't we just have built in dynamic range compression metadata like film soundtracks have? That way one could have stronger DRC when listening to music in the car, or over a crappy Bluetooth speaker, where it's actually useful, and less or no compression when listening at home. The way compression is used in most modern recordings caters to the assumption listeners will be using things like car stereos or speakers that just can't properly represent the full dynamic range. Movies have a massive amount of dynamic range, but there are plenty of people who watch movies on crappy laptop speakers or TV speakers with no issue, since compression can be applied on the fly.
Sent from my 5065N using Tapatalk