There are some dark corners of this crazy universe where adding arms and glyphs saves area.  And it seems this puzzle is one such.  

After building a stake-in-the-ground 79a, I submitted a solid 78a + 1400c machine that only contained the rightmost mechanisms.  As most polymer solves do, it advanced the growing product by sliding.  I was fairly pleased with it.

Noticing that the secondary metric was C and not G, I casually added bonders and arms in the empty areas of the product silhouette, to reduce the burden on Arm 1 and advance monomers more quickly down the chain...only to realize that half of those arms weren't even necessary.  And the hexes they occupied could go fully unused!  

So 78a became 75a, then 74.  By adding more machinery and cutting cycles by 36%. Consummately logical.

What a world we live in...