HK, to answer your question more fully beyond my edits, I think one key thing I do that isn't readily evident in course outlines is I draw feedback arrows from each input topic into other topics. Before I get started, I like to know loosely what causes what to have a "full equilibrium" model of the topic. Once I've wrapped my mind around each section, I'll usually do little thought experiments to stress test my initial hypotheses of how the sub-topics interact.
One diagrammatic tool I've found helpful instead of a simple flow chart is a series of concentric circles, kind of like valence shells, with topic labels next to points on the circles. It makes linking subjects easier and lets me think more generally about how topics interact. I'll usually colour code by concentric circle or by type of arrow.
Re: percolation, I definitely have a habit of tossing loose ends into my subconscious when I hit an impasse. Sometimes an overnight break does the trick, but I have occasionally needed months or years to overcome a hurdle. I recall a particularly vivid dream about grade 11 physics where the solution to a problem I'd been pounding away at appeared in big bubble font. I play Rock Band, the piano, or exercise heavily before exams or when learning something new, as those activities send my mind into the equivalent of neutral.
