The radiator before the fridge will do a better job than any practical sized resiviour in the fride. Fluid temp depends on rates of heat exchange, not with reserves.. To show this, really just add up the heat going into the coolant and the heat leaving (which is a bit harder to do... ), and the sign of that number will indicate which way the fluid temp will change. This can be proved using ordinary (second order? not sure
)differential equations. Notice that there is no dependance on the amount of fluid in the system... the amount of 'reserve' really just slows the change in temperature that ocurrs from the exchange of heat discussed above.
Argh, sorry to be so technical + abstract, but it's late and I like DE's.
Anyway... I'd dump the resivour idea, unless it's physically easy to implement (might be a good way to equalize pressure after the coolant reaches a 'chilled' equilibrium.. have an air trap in the top of the container to allow expansion / contraction.. hmm). But the best way would be to add a good radiator in the loop right before the fridge, so it takes the temp down as much as possible.
The other thing to watch out for is that dorm sized fridges don't have a lot of cooling 'horsepower', so it would almost be a necessity to put a radiator first.
Rich W.