Ok, so at 9,000 heat, your original design will output 40 cells off of the five inserted (as Akhkharu noted), so you'll have to "babysit" it at the 1/2 cycle point to do a cell-swap to get the most out of it (on top of regulating a 9,000 heat reactor, which is no trivial task).
The design I suggested will net you 80 cells per run if you do the swaps at the right times. (BTW, I didn't come up with it, I've seen it tossed around in some of the other CASUC breeder posts).
Basically, you get most breeding efficiency if you have every U cell touching 4 DU cells. Which leads me to the question why Akhkharu didn't set up his reactor with that in mind.