Here is their statement
I find this statement particularly unbelievable considering the difference we see in matrix metering in different bodies . My D40 meters very strongly for the focus point selected , the D90 is more accurate and I shoot in matrix metering with both of them .
Think about it, when a D4 is shooting at 10 frames per second is it really likely that in less than a tenth of a second the camera's processor can do all its calculations including flipping through 30 000 images before taking each picture ? Close to impossible I say !
It is more likely that Nikon have studied 30 000 images and written an algorithm based on that research to program matrix metering . There are many other factors involved in calculating exposure and each body is slightly different with different parameters programmed into it otherwise a D40 would give the same exposure as a D4 with the same scene - but this does not happen .
So my next lot of studies will involve analyzing matrix metering to see if I can work out the metering parameters of the D90 - I'm sure it will be similar to many other models out there and the more I understand a program the more benefit I can get from shooting in that mode by being able to 'predict' its reaction to different scenes at different distances.