This was by far the more fun metric of the two this week. The programming is so simple that a full redesign is not at all daunting, so I went through 4 or 5 of those; then one by one shaved 6 more points of area. Realising that the arm responsible for quicksilver could actually let go of it saved another 4 (before, I had that arm carry the product to the output alongside the hex arm). I think this is a very solid solution, but at the same time it feels like the top solves will be twice as small. Let's see!