I find it funny how swinging takes less area than pivoting if rotating the molecule fully.

I think area can maybe lower with some Berlo track shenanigans to utilize the area under the wheel, but with how many hexes are forced, I am unsure if geometry can enable this. 

My original area submission was an attempt for a double bonder 35 area that failed due to geometry at the end, becoming a 36. Two bonders removes the forced 3 hexes at the top due to not needing to rotate fully. Maybe that would allow a layout where the wheel can shave area.

Disjoint was fun to play with. Was useful to have the Berlo assign connected atoms the same cardinals while in progress to better distinguish them. This was then easily fixed once the shape has been assembled since Berlo rotations are infrequent enough to easily find.