boiled my first solution down, removed a bunch of track, area, and the very shameful arm. 

I think anything with a better rate would probably necessitate alternating between projecting and purifying the copper and then splitting those amongst the 2 outputs. Way above my pay grade, looking forward to the show.