I think my situation is quite different. I do very little coding actually too! I am quite independent and a purely theoretical physicist. I've already put out a

I think my situation is quite different. I do very little coding actually too! I am quite independent and a purely theoretical physicist. I've already put out a couple of papers and would be able to continue well without my supervisor. Although I have had really bad periods of procrastination throughout my PhD, but in general I work in short sharp bursts rather than sustained periods. You mention flow, it's hard to get into the flow when you are deriving new equations or approaches to things!