Man, am I gonna have to drag out the "expert studies" research again?

Plateauing like you described is human nature, more or less, and I think it mostly happens because we develop techniques and routines to a certain point, but never grow beyond them. I can't say I'm really an authority on the subject, and I'm certainly no role model, but as far as I understand, the key to surpassing a plateau is to consciously, deliberately seek to identify your limitations and weak points, then devise and test strategies for overcoming them.
You might find it helpful to listen to your finished music some time and try to identify specifically in what areas it feels deficient--perhaps write them down on paper (or on a computer, etc.) and then, having identified problematic areas, for each one think,"How can I improve this? Do I need to study more about [compression/mixing/instrumentation/harmony/etc.]? Are there video tutorials on this specific subject that might help? Can I post a topic about it on MLR?" If the problem is something that requires more "practice" or "raw skill development" rather than new knowledge
per se, ask, "Can I invent exercises for myself that will focus specifically on this weak point? Do I need to rethink my approach when I'm [designing synth patches/mastering/improvising/etc.]?"
After trying out a new approach or strategy for a while, reevaluate. "Did that take me closer or farther away from my goal? Do I need to be patient before I can expect to see results? Or should I just scrap that and try something else altogether?"
More generally, maybe a process like this:
1.
What needs to change? Identify problems very specifically. "It sucks" is not helpful. "It sucks because it has no energy" is better. "It has no energy because the rhythm section feels like it's dead or made of cheap plastic" is better still.
2.
How can I change it? Figure out ways to address the problem. I'd say there are probably three main options here, which may be useful separately or in combination with one another:
- Seek out help from others. Ask on forums, read what others have written, etc.
- Observe what others have done. Find examples of others' work that solves or side-steps the problem you identified in step 1, then try to analyze specifically what it is that makes them work.
- Devise your own methods. If you can't find any guidance from others or others' work, you may have to go it on your own. Maybe you'll need to try things somewhat randomly, hoping you'll stumble onto something successful. Maybe you already know roughly how to approach it, but you need to invent a more concrete exercise to be able to practice it. Maybe you need to do some more book-learning on the basics of whatever you're working on, and hope that something clicks, giving you the right idea of how to proceed.
3.
Is it working? If you're not fixing the problem, maybe you need to take a step further back and try to identify why your strategies or methods aren't helping you improve. Maybe you just need to give it more time. Maybe you're doing something wrong. Maybe it's an entirely bad idea to begin with.
TL;DR: Oh, just read it, you lazy bastards. Or take this as a summary: to improve, identify your limitations and then figure out how to overcome them. (Okay, that's basically a fortune cookie truism, but if you read the rest of my post you'll see some more detailed explanation.)
Now if you'll excuse me, I'm off to go ignore my own advice with a new project. :3