Here is something I hadn't noticed in the cycles diagrams before, and it's another reason why they're ... "suboptimal". That huge spike on the right side is actually the limit. Most people either solved this puzzle in 24 steps, which is the maximum number of symbols you can have, leading to a score of 2880 cycles, or they didn't solve it at all. If 2880 is the limit, why does the scale go up to 4000?
It is technically possible to finish a puzzle in more than 24 steps. You'd have to design it so that when the emitter hits the output instruction for the first time it's not pointing at anything, but the second time it is pointing at the finished molecule. But that's way complicated in a hard puzzle like this. If you can do that, surely you can finish it in less than 24 steps.
The final puzzle was much easier.