Often when I’m working through problems, I’ll find another one that looks good to solve.
It’s very akin to when you’re putting up a shelf – oh yeah, it needs to be level, I want it screwed into a stud, I want the screws hidden – and this then breeds things like – oh wait, the wall has a bend in it, I need to account for it, or the studs don’t line up as they should.
Some of these problems you need to fix now, others can wait (studs in the wrong place) – you have to choose what you change now, what gets updated, and what impact it will have on your current problem.
You can’t deconstruct the world all the time, otherwise nothing will get solved.