Thank you all! Is the color-correction masking what's described in "Photography - its materials and processes", p. 480-481? (available here https://archive.org/details/photographyitsma001646mbp)
"...the color corrections that most need to be made depend on the spectral absorptions of the dyes or pigments used in the color synthesis. As with other subtraetive colorants, the major adjustments to be made are due to the unsatisfactory green reflectance of the cyan colorant and the unsatisfactory blue reflectance of the cyan and magenta colorants".
In the case of carbro, I understand it so, that
1. because of the the unwanted green density in the cyan colorant, the negative for the magenta bromide (green-record negative) is masked with a positive made from the negative for the cyan bromide (red-record negative), and
2. because of the unwanted blue densities in the cyan and magenta colorants, the negative for yellow bromide (blue-record negative) is masked with a positive made from the negative for the cyan bromide (red-record negative) or magenta bromide (green-record negative) or "for better correction this mask can be made from a special negative exposed with a yellow analysis filter".
As for the pigment set, I suppose one could experiment with the pigments used by Keith Taylor, found in the article http://unblinkingeye.com/Articles/3CG2/3cg2.html (Daniel Smith watercolors Cadmium yellow light, Quinacridone rose and Pthalo blue red shade) or the pigments said to have been used in the Ultrastable process and by Todd Gangler: http://www.apug.org/forums/viewpost.php?p=1346515.