I am then going to work on a switch mode controller with a single control so that fully anti-clockwise the duty cycle is 100% green and 0% blue. Turning the control clockwise will increase the duty cycle of the blue and reduce that of the green so at the mid point they both receive 50% and fully clockwise it is 0% green and 100% blue. This will give me a single Contrast control.
The circuit I will use is one I have used before. It is almost a standard astable multivibrator but the base resistors to the two transistors are connected via a two diode and potentiometer circuit similar to the one in your link. There will then be a suitably large darlington transistor connected to either side to drive the green and blue channels.
If I built an LED enlarger, I would probably use a darlington transistor hooked up to the PWM outputs of a microcontroller, since I have microcontroller infrastructure already. Then you could control the R, G, B lights digitally as well as create automated recipes, or have feedback from a photosensor, creating a closed-loop exposure system.
That's quite a logical and sensible way of doing things as you can fine tune it to be whatever you want it to be and have settings to match proper grades etc.
However, I like to keep things simple myself so I'm going to go for my single contrast control!
So you're using the Darlington transistor so you can get more current out of the system to drive all those LEDs.
Whereas, my idea would be suitable for a couple-few LEDs at most and would peter out if I tried to drive a dozen or more of them, such as you plan to do. Right?
(P.S. - I read you were going to make a "controller" but I skimmed over the words "switch mode".)
Easy for me to say given that I know nothing about electronicsf, but I wish that an entrepreneurial sort here would devise 2 items:
A simple, adaptable LED variable contrast light source for 4x5 and 8x10
A simple Jobo alternative, something like an electric screwdriver turning a drum in a water bath.
Now that Aristo is moribund and the bulbs are currently out of production there could be a market for the light sources. Likewise with Jobo and everyone paying $500 for drums.
I wonder if something like a Beseler/Minolta 45A head could be converted to LEDs? It's already set up for RGB flash tubes... (which sadly are out of production and hard to find as replacement parts)