First week submitting, and sum4 looked extremely intimidating, with no clear direction to optimise into. After looking at some polymer sum4 solutions compared to sum, it seemed like on these it just meant "don't 6p it" which was entirely fine by me, though that necessitated building the output backwards. The only thing I was fairly certain of going into this was wanting a length 2 hexarm to play with the glyph of division, because being able to grab both irons at once felt really efficient, and that did help to constrain decisions elsewhere, but everything else felt rather up in the air. At this point though, I'm unsure whether that arm is the only good thing, or the root of all evil, as the swings have been quite annoying. The main thing I wish I could resolve is needing a dedicated arm for each of mors/gold/waste, geometry means most everywhere I could put the mors for the gold to grab it efficiently will have the iron hitting it, going back to grabbing the input by the gold arm rather than the waste one would also be nice to save a cycle but that would be difficult. And of course outputting the second half-output feels terrible, but there's no way to make it better without adding another dedicated arm, trying to get the mors arm to swing the output does fix the rotation issue but the output arm doesn't wind up holding it by the left iron atom, which causes problems. I'm assuming min will be something like 350, but I lack the experience to say for sure.