Here's my 2 cents worth.
Linotype-Hell's, (presumably now defunct) foray into extended gamut separations called "Hi-Fi Color" was essentially the same as Gordo described for the Hexachrome technique, except that Hi-Fi color was setup to create a maximum of 7 colors, CMYK + RGB.
The intended purpose of the extended gamut was to address the RGB deficiencies inherent in the CMYK color space. The RGB deficiencies in the CMYK color space I suspect we can all agree, are not along the neutral axis. So in the case of Hi-Fi color, and I presume for Hexachrome as well, both techniques were purposely designed so that they did not extend into, nor interfere with the neutral axis.
As I understand it(leaving UCR/GCR aside for the moment), the maximum TAC value in CMYK printing occurs in the darkest shadows on the neutral axis and then tapers off from there. Extended gamut separations, never being designed to extend into the neutral axis as mentioned above, therefore should have no influence on TAC values.
The same technique described below by Gordo for Hexachrome, was also employed with Hi-Fi color.
With Hexachrome because process color is removed where the extra colors are used during the separation process. I.e. Where orange prints, a percent of yellow and magenta are removed. Where green prints, a percent of yellow and cyan are removed.
So this leaves me with a question regarding ProfileMaker. Does ProfileMaker make any distinction between an extended gamut job and an ordinary 6 or 7 color job? It seems to me that it should.
Best Regards
Otherthoughts