A Few More Items:
First, my statement that "The earliest dye-transfer (or dye imbibition) prints were made by exploiting the same mechanism that modern day carbon & gum printers use" is possibly true, but slightly misleading. I think it's fair to say that silver-halide matrix films were utilized just as early.
Another important issue of 3-color assembly printing is registration. The typical answer to this question is to say, "go get some Condit Mfg. punches, film carriers and registration boards." Your response should then be, "honey, how about a 3rd mortgage?". Thanks, thank you... I'll be here all night...
But seriously, if and when such items appear on eBay, the prices are exorbitant. Not a sustainable solution!
F.E. Ives deals specifically with the issue of registration in his patent #1,121,187 (linked above). He proposes two solutions; 1 is to align visually, which he calls a perfectly satisfactory method. The novelty of his patent claim though, is in the use of a "separator" between the matrix and print to delay imbibition until registration is secured. This "separator" is nothing more than a thin layer of acidified water that is to be removed by going through a roller-press or squeegeed out. As you'll recall, an acid solution will keep the dye in the gelatin.
His 2nd solution is the more familiar registration board. It doesn't consist of pins, but rather you simply align the matrices, clip one edge so that they are all flush, and place it on a board with a corresponding edge to press the matrices against. Pretty basic...
My goal/plan is to use a simple 2-hole office punch and a simple registration board. Assuming we're at the point where we have 3 enlarged separation negatives, they must be registered visually on a light table. A special mark or insignia (like a star, asterisk, etc.) on the negatives will be very useful for this. Once they're registered, they are punched together. Likewise, the matrix blanks are punched before exposure, and each matrix & negative combo is placed in registration during exposure. Then, the final transfers are made on the same board.
Alternatively, if you don't want to punch your precious negatives (probably not a bad line of thought...), you could use a system where you tape another piece of plastic to the negative and punch that. I believe Andrea Zalme does something like this for her 3-color carbon stuff.
The other option is to align the dyed matrices in a 10% acetic acid bath, and just punch the matrices at that point.
So, in theory this should work well. The only special piece of equipment will be a home-made registration board. There are a number of manufacturers who sell custom-sized registration pins. These can be easily nailed (some have a hole in the center) or glued to a good piece of wood, MDF, etc., using a punched piece of film as a guide. Or, you can go to the hardware store and you'll be amazed at how many things might make good pins; these could be spun on a drill and sand-papered down if they're a little to big.