[old_tv_nut]

Narrow cut filters are one technique that has been used in color separation to increase saturation since the beginning of color graphics art. Masking using a negative of one color separation to modify one of the others was a common technique to compensate for the deficiencies in printing inks. (The magenta ink was usually the worst, having far too little blue reflectance.) The amount of masking could be reduced over the years as printing techniques and inks evolved closer to pure primary cyan, yellow, and magenta.

Technicolor depended both on sharp cut filters and printing to high contrast to enhance saturation, but provided no inter-separation masking. When they went from filters to dichroic mirrors in the late 1940s, the bandwidths were much greater, but the crossover regions were very steep.

Multilayer film has added inter-layer inhibition - the products of development in one layer inhibit development in an adjacent layer, thus enhancing the difference between them. I know Kodak has used this in some versions of color negative material in the past to produce a high-saturation version of a product range, and I'd guess probably uses it in Ektar. I wonder if what you describe is due to unwanted variation in the strength of interlayer interaction. This seems to align with the observation that it is batch related.