Kinda like where that idea in the picture is going.
I could make my own suggestion thread about this but i doubt it would ever get implemented so...
Instead of having chambers etc etc, make uranium rods that can be controlled by a piston like block to push/pull them.
What you do is you push them into a containment chamber (A bunch of blocks that you have placed) and they start reacting with other uranium rods.
Then you have a new kind of liquid (that is finite) that we can simply call reactor liquid. If the uranium rods heat up (due to them reacting with each other) the reactor liquid is evaporated.
This evaporation event is picked up by a steam collector (It just knows a reactor liquid evaporated because it was informed of the event) which will produce EU's for every evaporated RL that "passes" through it. (Maybe make it more efficient the higher the energy of the evaporation event is)
If there are more "steam" being produced that a single steam collector can handle then the evaporated RL will instead turn into a fog block. This is BAD because this fog block cannot be replaced with a coolant block. If you start getting fog blocks then your reactor will run out of control unless you pull the rods out (a slow process perhaps?)
Since RL is finite, you have to produce it using IC machinery and feed it into one or more coolant injectors. These will replace any air block inside the containment with RL, which cools the uranium rods and turns to reactor steam.
Any evaporated RL that goes into the steam collector / generator is either stored in the collector (to a certain point where it cannot store anymore) and to extract it you need to feed the collector with empty containers (a special type of item designed to contain the radioactiveness of the RL that has been through the reactor.
To re-use RL you have to process it, you get one part clean RL and one part reactor waste (which can be used in the nuke recipe or used to heat up breeders.
Well, you probably get the basics of this idea by now. Make the reactor a 3D experience, using lightweight code to implement it. Cache everything and only update that which has been changed when it changes. A complicated process but worth it IMHO.
Such a reactor will also need to have it's update interval reduced. But it doesn't matter much though, the steam collector /generator can still use collected steam as "fuel" and output a steady flow of EU's.
So far, this only implements the SUC/CASUC part of reactors. As for self contained reactors, you can add heat sinks in the reactor core to reduce the overall heat in the reactor. Reactor Liquid will be formed slower and any fog that would appear is taken up by heat sinks and re-converted to reactor liquid. Of course having heat sinks will reduce the rate of evaporation in the reactor chamber but that is needed unless you want to feed your reactor with fresh reactor liquid throughout the process.
So, how does this design differ so much from the GUI version we have now? Well, the longer the rods, the better they perform but you also need to have a good mixture of reactor liquids and heat sinks etc.
It's equally important to balance the energy generation with energy consumption in a reactor design, certain components will draw precious EU's to operate. Creating RL is a costly process in both materials and energy but with an efficient reactor design this will still give an overall gain of EU's.
I leave the technical details/math unexplored right now since like i said, i doubt this will ever happen. But if it was to be implemented i would be willing to help out with that.
