This is a pretty simple metric to create theory for, at least compared to the others we've been seeing lately. Clearly 2 width is not enough to move the input anywhere, so 2.5 is min. Also, the input, output, and unification glyph all cannot be reached over by an arm (nor moved past), leading to a rather trivial need for at least 2 arms, as no individual arm can reach everything needed. Additionally, no arm can reach every tile of unification, so that becomes the center of the solve, with the 2 arms on either side. Additionally, it's impossible to take the full input to a debonder (or similarly, take the full product to the output from a debonder) with and less than a piston arm on 2 tracks, thus giving us our minimum cost components being a bonder, debonder, unification, 2 pistons, and 4 track, totaling 160g.

I have no idea what cycles look like here though. this is probably bad in that respect.