Whenever I need to adjust levels withing a DAW track...I prefer to slice up the track into smaller pieces, the size of which is usually governed by the track's content (like the individual words of a vocal track)...and then I just raise/lower the level of that sound wave piece
One of my associates works that exact same way, and has gotten good at it and fast at it. It's a valid style.
Just like DAW software selection itslef, it all depends on what one gets used to doing and using. I have no problems with other people doing things that way, but personally I prefer the rubber bands for a few reasons.
First, what you call "clutter", I call "informative". I like seeing the gain and pan automation curves as I can tell at a glance what's going on.
Second, gain automation is simply the digital version of the original riding the faders in analog, which is the way God originally intended it (
). Chopping track up into segments and treating them as individual objects with individual volume settings just seems artificial and counter-intuitive to me.
Third, and probably the biggest, is when you want to do do more than just general quantum jumps in volume - i.e. ramp the volume over time, you're going to want to automate that anyway. As long as you're going to have an automation curve, you might as well use it as the common control instead of using one in this segment, not using one in that segment, but using one in that other segment, etc. and then having to track where you're using it and where you're not.
The exception to all this for me is VST automation. I have no use for that to begin with (who the hell really needs to automate a VST?), and when you throw too many rubber bands on the screen, then I agree it's just clutter. But for gain and pan, I'm making the Koolaid
G.