While impressive that the optimal can be proven, I feel like the example puzzles here aren't ones that are particularly hard to find solutions for (when move count doesnt matter). I'd be interested to see at least one example that has a lot of tricky dead states that would act as traps.