It can be difficult to find new mechanics in the programming game space, but I think that this demonstrates that there's still plenty left to mine.
I was concerned after the first two puzzles that there wouldn't be enough to build a game a from, but smart level design puzzles proved that you can build robust problems from a simple toolset. I particularly liked the 'reset' mechanic between zones.
The way that I judge a puzzle game is how it makes you feel -- the goal being to have a routine answer make the player feel like they just solved quantum mechanics. And to that end I think you definitely succeeded.