Always better to nail one thing down really well before you start thinking of other possibilities for a project/product, because it can most certainly distract you from accomplishing the small wins. Dwelling too much on all the things you can do over what you need to do at that very moment has proven to be toxic time and time again. As someone who has committed this very crime, I definitely understand where you're coming from.

I can not tell you the countless hours (which turn into days), that I've spent trying to perfectly code or execute a design or method. I've learned that the best steps to completion that I use (this is in regard to software and website dev)...

1) think of the idea

2) code it (notate any epiphanies along the way)

3) optimize the code

4) package and ship