Sounds like you are working on deep complex systems - i totally get the experience of needing a couple of hours to stare at how the system behaves before the li

Sounds like you are working on deep complex systems - i totally get the experience of needing a couple of hours to stare at how the system behaves before the lightbulb moment when you realise how a simple change will achieve what you want (ideally!). There are other times when it's a bit more greenfield and just fiddling around gets your ideas straight and you can be productive even with interruptions. We tend to work with a backlog of simpler and harder problems, and part of the skill of prioritisation is deciding what mood you are in, and being realistic about how much time you have to focus on a problem, and hence what sort of problem or feature is worth tackling that day.