I'm shooting some 8x10" Polaroids under window light, so I had another chance to test the meters, and I threw my simple Gossen Sixticolor into the mix, which only reads red/blue, so it can't handle fluorescents, and it doesn't work terribly well in low light, so it's not that great for tungsten.
Well the first shot revealed that I needed a warming filter, so it seemed like a good situation to compare the three meters, though of course I could have made the correction from looking at the Polaroid itself.
First off, the Minolta IIIF is more sensitive than the others. I could get a reasonable reading from the subject position, maybe 8 feet from the window. The Minolta II and Sixticolor required that I go up to the window, and it could be that the light is bluer 8 feet away than it is right there, due to reflections in the room.
All three gave me about the same reading calling for something in the range of 81, 81A, or 81B, and it could be different depending on how precisely I directed the sensor. The conclusion might be that there is no meter that can tell you precisely whether you need an 81, 81A, or 81B, because there are just too many variables involved in the process, but it does seem informative to have some readings to base a judgment on.
As to the question of why one might want to use a color meter, I think they get the most use in mixed lighting where you have the possibility of gelling all the lights to the same color balance, or making multiple exposures with different filtration on the lens for each light source, as is often done for architectural interior photography.