Yeah, it really depends on usecase you think about. If you speak about compession effecting something like an electronic kickdrum then yeah, you can draw automation instead of it. (When speaking about clean non-saturating compressors) ... but the advantage of compressor is that on stuff with "unpredictable" and/or quick transients (real drums, vocals, pianos, synths with a lot of modulation etc...) it will do the trick for you automatically.

So theoretically you could draw "compression-like" automation over a master channel or finished track, but it would take you weeks or months to do it. Compressor will do that for you instantly.

I have a similar feeling. It's good thing to have, it's more "professional" way and the difference is definitely there. ...but based on how delicate the change is, I'd say this investment is not that impotant if one's on budget.

Here's a track to set the topic's mood: https://www.youtube.com/watch?v=v2H4l9RpkwM

We all have them. We all fight them. So let's start a little group therapy here:

I fight with several bad habits:
- I have a big fear that I might upload something that I like, but it's actually sh***y. Happened before, wasn't nice, thus the fear.
- I have this urge to make a prominent melody even when it's not needed, or when it's actually a bad idea.
- I forget about how important little details are.
- I press play and try to come up with next step when the song is playing. That makes over-listening comming really quick every time.
- I don't give up on projects that even kill me to listen to. Big one.
- I tend to do mix adjustments after the track was uploaded to soundcloud.
- I do prety drastic musical changes really late in production process. That makes re-sampling totally irelevant technique for me.
...and more.

What's yours?

Yes, I think that this problem is all too common with beginning producers (and even some very experienced ones) who spend way too much time trying make their mix sound "pro", when the 7-minute song itself has nothing musically interesting in it to keep my attention for more than 20 seconds.  Songs that go nowhere, have no melody, and, hell, often not even a chord change, but, the mix actually sounds pretty good - I find this to be the case more often than not on production forums.  IMO, the primary purpose of the production part of music-making is to bring out the musical ideas, but too many people think that they can get away with skipping the hard part of creating great musical ideas simply by having a stellar production.  IOW, yes, you certainly can polish a musical turd, but it just leaves you with a shinier (overproduced) turd.

Agreed with every single word.

That's actually the point. I don't. ...but it feels wrong. It feels wrong to just suppose there isn't such a thing in electronic music. When listening to some IDM stuff I sometimes think that some of the stuff they do is too complex for my taste, but probably wasn't for theirs.

Theoretically do you think there is a complexity level for a track where it would become evident that producer crossed some line?

With music that is based on some kind of instrumental or vocal performance, it's quite simple. It's overproduced when the performance is not the main thing anymore.
But the same question is much more interesting in electronic music world. That reference point is gone, so everyone has to invent their own. What's yours? How do you find out that you're trying too much?

I suppose the main answer will be: "It's overproducing when I'm hurting the song." ...but that's too vague. Where do you look exactly to evaluate this?

For me, It's usually when the feeling of the very first thing that "made the song" is starting the disappear. When that piece of melody or rythm that made me continue with the track is getting away. That's when I start to be a bit worried about overproducing the track. You?

Bertie said it all. It's good tool for ocasional checks, but nothing to base descisions on.

Yea, but some of the best sounding stuff to me, and arguably many others, are when the drums are causing the peaks.

Yeah, I ment the extreme peaks. Thus the number. It's always good to leave some dynamics there.