1/4 + 1/20 on both outputs

My wild guess of a prediction is that winner will be around 500G and that this solution will place around 20th. (The competition do be getting stiffer and stiffer each tourney!!)

I started with the left hand side which I'm reasonably happy with, then attempted to build around that. This led to eventual monstrosity of the right hand side after numerous painful hours.

I went with the first algorithm that came to mind - There's almost certainly better out there. Dual Pipeline is probably not the way to go here with how much of a nightmare input routing is - I would guess that Brick runs supreme here.

Potential Local improvements:
Arms 16,17,18 could be combined with better geometry. 
Arm 13 is very sad and could be combined with 14 and/or 20, although getting to re-use the calcifier is nice.
Could probably re-use a single bonder on the left and in the middle, instead of having two.