I know that Eizo makes their comparisons using the equivalent of a 2D gamut projection. So this would be based on essentially a 2D outline of the gamuts. I’m pretty sure other manufacturers do the same.
As those familiar with ColorThink know, a 2D comparison of gamuts does not tell the whole story. There can be many portions of a seemingly small gamut that exceed those of a seemingly larger gamut (especially in shadow areas).
Also, when comparing gamuts, keep in mind the difference between coverage and size. A model might have 114% of the size of NTSC, but does it actually cover the same area (same colors) as the NTSC gamut? Any comparison between numbers like these should include size and coverage if you are really depending on this monitor to give you the gamut of something like AdobeRGB for example.
BTW, 3D gamut calculation like what’s done in ColorThink Pro is not so easy. We have always said that it is a coarse estimate. To be able to “wrap a skin” around a 3D gamut and calculate the volume of it is computationally expensive, and even more so the ‘finer’ you try to get. It works best on devices that have smooth transitions (like monitor profiles) and will be less precise on irregular printer profiles.