I would have to run the numbers to be sure but; I think if you upgrade your coolant cells to their biggest size and then run them in a freezer then it should lower your overhead running cost.
I actually started out with 360k Helium coolant cells. The vacuum freezer scales differently for water/helium/nak cells, but overall the time used is:
2 kHeat/Second (Water cells), 2.4 kHeat/Second (NaK cells) and 1.71/1.80/1.85 kHeat/Second (60/180/360K Helium cells). So larger cells don't really help that much, if you're using a freezer. But larger cells would make me need to cycle the cells to the cooling tower less often, which would indeed save some power. I mostly used the smaller 60k cells because I didn't have enough Potassium or Helium.
I've been unable to come up with a reliable way of removing and replacing the cells correctly.
I use Factorization routers for inserting fuel cells, and removing coolant cells that accidentally get placed in the fuel slot when the plutonium runs out. Routers can selectively input or output to specific inventory slots in reactors, which is very handy (slots 0-53, going left to right, then top to bottom, treating reactors with less than 6 chambers as full sized reactors, with the extra slots piled on top of each other in the right-hand column). They can get expensive if you need item filters, though, plus you need a different router for every slot.
Omicron: I agree entirely about using Thorium reactors in parallel. I have 10 set up in my world right now, using:
http://www.talonfiremage.pwp.b…uii2mi6vc00eq1tcgc546x1j4
Which only has an efficiency of 5.67 (and probably is over-engineered on the cooling, due to the aforementioned discrepancies with gregtech), and uses single cells as pseudo-neutron reflectors. The reasons I didn't use Thorium reflectors for the plutonium design were that first, I had the iridium on hand due to quarrying out huge areas of a mystcraft age, second, the extra heat would have actually been slightly more than my design could handle (9.6 heat being greater than the 4 extra I have), though I could have handled that by replacing the component heat vents with reactor heat vents, and lastly I just liked the idea of a 'perfect' efficiency plutonium reactor.