I had an opportunity to do extensive uniformity testing on a LaCie 324 early on. Our test model would wander a bit in it in terms of color consistency from side to side, and from corner to corner around the screen. The average delta E difference between different points on the screen averaged about 2-3 delta E, with the worst points being 6.8 dE (from upper right to lower left). (Delta E is a measure of how far one color is from another. Anything greater than 1.0 is visible with the human eye.) This sounds like a lot of color variation, but in actual use a display with these numbers will generally look pretty uniform, although you would be able to detect some color shift when viewing a perfectly gray background. These numbers are about average for a display in this class. To get something better, you would want to look at an Eizo CG222W. They pretty much guarantee that uniformity across the screen is less than 3.00 dE.
When you talk about an extreme shift to the green I don’t know how much is extreme for you. If you have the ability to take spot measurements, you could compare the different corners of your display to each other and see how far apart (color-wise) they are.
At any rate, it looks like you would want to profile your display by putting the sensor in the middle of the screen. There’s not really anything else that can be done to compensate for a screen with bad uniformity. If you got this model for testing, can you return it?