Based on the theme, was expecting the metrics to spell out words for a joke message, but that would probably leave them unfun to optimize. Wonder how many people were as paranoid as me about the name changing the score, so submitted their first solve under different names.

Found it fun finding the exact rate needed, after being shocked when a rate 1 solve was not as low as expected.

Was interesting to have a forced rate that was not min. That gave more wiggle room in cost allocation, which allowed a single hex-arm to handle the input instead of multiple arms. 

There is still some variance on this type of layout, which makes me think better placement could remove arm 5, reaching 150G.