To me it seems to progress:

1. You just wrote the music - it sounds great!

2. Next day, you're not so sure. Maybe some changes?

3. Perhaps if I destroy some parts. Salvage Job.

4. This music sucks no matter what I do to it. And using compression just ruined it.

5. Oh that one, I can't be bothered with it anymore.

6. Well adding that seemed to have helped a lot. It'll have to do.

7. Hey, I don't mind that at all!

I vaguely think there's two kinds of music - that which is finished at stage 1 or 2 and that which is finished at 6. Bradbury for example would rather spend 4 years to get to the end than give up. I see a lot of benefit in stopping at 2 or 3, because often that first enthusiasm is true for the listener.

Is the music for your own entertainment? Or if it is for a purpose? In commercial music production you might be sent back to do it again and again - it's the job - and you know that you are trying to hit a goal. Do you have declared goal? This track is supposed to achieve -this-. For example, the Music Server series had to be played at my workplace without the other people complaining, or even noticing.

Put yourself in the hands of another. Electronic music is too often a solo process, either let someone else change it (I used to let Racic ruin everything) or let a process change it. Brian Eno was always on about this. First of all he has his Oblique Strategy cards - you had to obey them. And he has rules like - not adding new instruments, rather treating those that are there. Once you add an instrument you must not remove it - only mix it such that it does no harm.