Not a whole lot of progress has been made, but here are my latest offerings to share.
For this round of experimentation I took a matrix-film that has had an extinction dye added and spirit sensitized with 90% isopropyl alcohol & 5% potassium dichromate (1:1, IIRC.. it's written down somewhere). The composition of the coating is 6% knox gelatin, 5% McCormicks yellow food coloring (primarily tartrazine, or yellow 5) and 1% sorbitol, coated with a comb on PF subbed melinex.
The 1st scan is a transfer made to Kodak DT paper for 10 minutes with Kodak Magenta dye/buffer; with a prior rinse in 2% acetic acid. The 2nd scan is the dyed matrix that has been allowed to dry out, leaving a magenta transparency.
My biggest problem is this damned splotchiness, which appears to be portions where gelatin has been removed. I haven't quite figured out the problem yet, and this seems to be the worst example so far, so I don't know if my new sensitizing method didn't go so well, or if my etching procedure is giving trouble.
Also, it appears that the matrix is way too high in contrast, and I probably need to increase my dichromate amount, which might get tricky with potassium dichromate.
I'm wondering if any carbon printers have seen similar patterns, or if any dye-transfer printers recognize it either.
At any rate, last night I coated 7 matrix sheets (largest batch to date), and I intend to do some woodshedding to get to the bottom of this.