I spent a lot of time mixing where I didn't have the benefit of everything being automated. Sounds like you have too. Before I got Cakewalk, I spent a number of years mixing music using all in one solutions like the Akai DPS 12 and Tascam 2488. With both of these recorders I would make live recordings and then mix them again and again for a real-time bounce down where I was sliding faders etc. Sometimes when I listen back to those old recordings I still prefer them.
Cakewalk makes me way more efficient. I mix substantially more music in substantially less time, but sometimes the speed shows.
And sometimes, the more I listen the worse I make a tune. It really depends. Less is often more.
But I love experimenting where I have no idea what I'm doing. I have certain friends who I record and I pretty much know the FX I'll use. I know they don't love it when I apply a reverse reverb (giggle) or a reverse delay (ha ha).
I try to keep experimenting.
At this point, I definitely mix anything I care about using Sonar.
Frankly, it's pretty easy to level songs for a CD. Find the quietest song for the CD normalized and then lower the volume of other songs relative to that one so that the listener doesn't need to adjust the volume while they listen to the CD. You can get fancier by remixing the louder songs so they contain more dynamic range rather than just reducing the volume. When I say reducing the volume, I mean lowering the volume fader so that the tune doesn't use every available bit. If you increase the dynamic range, there can be a few loud moments in the song (less compression) but the song itself has a listening volume closer to the softer song while still using every available bit.