from biggie:
my speedsolve used this 2-atom input.  I knew it would be the simplest for sum because it works wastelessly, and I declared early on that it was the only correct input for sum.  I made a few solutions with 2 arms trying to target a 16 or 14 cycle loop, and my best was 252 at 105/110/37.  It had two debonders and one of the arms was on a triangle track.

Then I decided to rule out wasteful solutions by building one with a 3-atom input.  Using the bent 120 degree salt-air-vitae that features in the CG solve, I got to 256.  Then using the straight line salt-air-vitae I got a 247 at 110/77/60.  The assembly in this was slick with 12 cycle loop and low glyph count.

It took until the last day for me to go back and use the same assembly that featured in that 247, on a solution using the 2-atom input.  Immediately I made a 235, and gradually cut it down to 215.  So it was good intuition that this was the only correct input for sum, it just was a rocky road to get there!

All iterations were shared in the Team 42 DM with morra and genius42.  Morra as a playtester is just watching along for enjoyment's sake, genius offered some testing and advice along the way.