If you track your pace hiking one thing you discover is that a non-trivial break almost never pays off (presuming you're past a certain baseline fitness level).

If you track your pace hiking one thing you discover is that a non-trivial break almost never pays off (presuming you're past a certain baseline fitness level). Unless you get lost. I recently discovered I spent days cranking out unpleasant repetitive code when I should have used a different data model that would have allowed me to write a trivial generic solution. I suppose this is an area where no heuristic is anywhere as good as having the experience to know the right answer in your specific situation.