I'm not even sure why it's showing you differences based on RI. ColorThink is doing this correctly IMHO as explained below. It's the only gamut viewer I'm aware of (at least on Mac) that is doing this correctly. Not the CS utility, not X-rite's offerings.
I don't know ColorThink at all so I can't comment on what it does.
GamutVision allows you to view the gamut in a number of different ways (and also density response curves and other things). If you want to see the gamut as it is, you need to either select the profile as an Input profile, or if it's selected as an Output profile the Rendering Intent should be None (in which case you can see the full gamuts of the both profiles at the same time, which is useful for comparing gamuts).
The reason for being able to select a rendering intent is that it will show the mapping from Input to Output. So if the Input is Adobe RGB and the Output is a printer profile, the different RIs will show the gamut clipping, direction of the clipping, direction of the shifting of colors etc. Swapping Input and Output shows the reverse process. All very useful.
I would be surprised if GamutVision is doing things wrong, seeing as how Imatest is pretty big in this business, but then software is software and bugs are always possible.
Robert