Hi Bob, I will definitely give this a shot...
Keep in mind, I have actually never made a color carbon print, or even a dye-transfer print, although I speak as an authority on both topics. I'm only an authority in the sense that I'm well read on the topics and am taking steps to make these prints in the future. Plus, they fascinate me! At the moment I kind of consider myself a "pied piper" of these processes and approach them from an academic standpoint. But, photoraphy as art is where my interest lies and where my aims point, and it is with this in mind that I pursue these processes. If only there were more time in the day where I'm not stuck behind a desk...
In mechanics, creating a color carbon print is no different than creating a monochrome carbon print, with the exception of registration. Numerous graphic arts punches have been made throughout the years, and these are all quite expensive. My goal is to use simple office punches, and I have reason to believe that these will be adequate.
The first thing to do is obtain the pigments that are necessary for color reproduction; cyan, magenta and yellow. I got a pint of each of these pigment powders from a place called Lansco Colors, for free as samples.
Here's a post describing the pigments, and here are the ones I got from Lansco. Cyan, Magenta, Yellow. The cyan and magenta are the proper pigments, the yellow was more of an educated guess but I would guess that it should work ok. APUG member mdm has suggested using painter's tints which are conveniently supplied as liquid dispersions.
The tissues are made just like monochrome tissues, but the trick would be in calibrating the 3 colors to create a neutral gray scale when overlayed. Chances are that this would entail a balancing of pigment concentration, gelatin concentration and sensitizer concentration. Furthermore, since carbon relies on UV light and the different colors absorb/reflect this light to different degrees, matching the contrasts is further complicated. This wouldn't be a problem with carbro. Luis Nadeau specifically addresses this and mentions that is most pronounced with cyan.
It sounds daunting, but one place to start would be to make small batches of "glop" containing the 3 pigments at various concentrations. So, let's say you make 10, 20 and 30 (totally arbitrary numbers) grams/liter of each color pigment in 10% gelatin. Now pour these over one another in all the possible combinations. The combination that creates the most neutral gray might be a good place to start regarding pigment concentration. Then, sensitizer affects could be investigated to bring the 3 colors to the proper contrast and scale.
Finding the proper proportions for each color tissue would be the most important and intensive step in a workable color-carbon scheme, methinks. But frankly, it sounds kinda fun!
Now, you need to consider your color separation negatives. This is where my expertise is lacking, but fortunately this is where there is a plethora of information. From classic analog controls via developers and masking to QTR profiles and digital inkjet negatives, there's no shortage of knowledgable people on this topic. Dye-transfer would be a good place to look for sensitometric information and methods for analog separations and masking.
Color corrective masking would be important in an analog scheme and equally important but trivial in terms of effort with digital.
At this point you've got your negatives and your tissues. Assuming double-transfer, you will find yourself with a suitable temporary support be it dichromated albumen coated on mylar (excellent document explaining this on the Yahoo! Carbon Transfer group) or just plain melinex that has been washed and scrubbed with Comet (as per the UltraStable instructions). You'll want to punch this support and you'll need a registration board with pins that matches your punch to hold the temp-support and tissues throughout development. The board might be the only piece of specialized equipment you need, besides maybe a Meyer rod for coating.
Yellow is the most opaque pigment, so this will be transfered to the temporary support last, placing it at the bottom of the final support. There is an excellent video posted by Charles of Tod G. doing this whole process.
This is a pretty basic explanation, and perhaps I've left out some crucial information (?) but in reality I think that anyone could tackle this, and the more people that do it the more the collective knowledge will grow. All we need is a starting point.