A very similar experience here (although not to quite the same level of success!).

Even for simpler problems the ability to just sit with it for a few hours/days dramatically increases your ability to write quality solutions.

Good code isn’t the most difficult or complex CS algorithms. It comes from thinking about the problem until you understand it so well that the code is almost self-evident.

Of course, once you’ve thought things through so much it seems trivial to you - which I blame for the annoying verbal tick we all tend to get when explaining things where we say “Basically it’s…”, or “it’s really quite simple…”

The real fun starts when you get to the problems that _do_ need weeks of thinking, and multiple prototypes to validate your theories. That’s when you start to feel like the “science” in “computer science” might not be misnomer :)