For your answer for the second question, I would argue that long range dependence is what makes for a hard puzzle, which is often considered rewarding, because, even if it's not planed, limiting the solution space allows for more complex logic.
