The theory min area solution juggles 12 waste.  It takes 19 area to do this while also preventing accidental bonds.  This is because you must move one thing at a time.  Moving more than one thing at a time only starts to buy you a true area savings when you can move a set of atoms or molecules in a complete cycle, which by my estimation requires more than 2 arms.

So while I expect the winning area solve is 19-1 = amarms of 18, the true amarms theory min is actually 17.

I was able to make the final output at 17 amarms in a test puzzle with scaffolding waste.  Watch from cycle 208 to the end to see that here, which I would call the "actually optimized" portion of this solve.

To make the rest of this, I made a faster program with the same arms (some upgraded to pistons) for outputs 1 and 2, and then just used the final output programming for outputs 3-6.  I'm sure that if all 6 outputs were solved as optimally as possible, including adding arms in other places, it could go faster.

Genius42 suggested "the pacemaker" as a solution title here, because of the resemblance to a heart.