Capstaff in his patents, see US 1315464 issued 1919, calls for "an acid dye (preferably a salt of a sulfonic acid)". I take this to mean the molecule has SO3 groups. These SO3 groups would cause the dye to bind with the gelatin. A wide variety of dyes were available to him at that time and probably there was one of the proper shade of orange and one in the proper shade of green. However, it is certainly possible that the 1920's Kodachrome used a mixture of dyes for either the orange or the green or both.
Dye transfer processes often used a mixture of dyes to form one of the fundamental shades. Technicolor mixed its dyes both to form the cyan and the magenta, IIRC.
Cinecolor, although not DT, famously used a mixture in its two color process for the red tones. The blue tone would be a silver image on one side of the film which was converted to a cyanotype (Prussian blue) by ferric cyanide. But the red would consist of a maroon dye and an orange dye which attached to a mordant created out of the silver image. Red objects would look natural. However when a yellow object was photographed it left a less dense silver deposit in the image and only the orange dye adhered to that. Thus yellow objects appeared orange. The two dyes would indeed express in unequal ratios depending on the exposure, but it had the effect of producing a separate tone for the yellows. Some people said that Cinecolor was a two and a half color process instead of three.
All the DT processes offered considerable post-exposure latitude to the lab. While the results might not meet the high standards of the experts at EK, they are a form of art in their own right, IMHO.