Figured i'd try my hand at the sum puzzle after having burnt out, so have an afternoon's effort. 

New Solution 1 was a 325G/90C 1/2P6 solve clocking in at 631 Sum4: I then made that cheaper and slower.

Arms 2 & 3 work at pseudoperiod 8 (16R) but are limited by Arms 1 & 4 at pseudoperiod 12 (24R). This means I'm leaving a potential 8R (40 cycles) on the table. Thus, I predict the winner is sub-500 Sum4.