Numerical values in TIFF files?

Y! Answers isn't allowing me to add details to my question, so I'll do it here:

If I create differing levels of gray and examine numerical values contained in a 16-bit RGB TIFF file, black is represented as 0, white is 65535 and 50% gray is 32767. The same value exists in all three channels. If I create differing levels of gray, the value is simply FractionWhite*65536 - 1, where FractionWhite is expressed as a decimal between 0.0 and 1.0, with 1.0 representing pure white

If I create a 16-bit gray scale TIFF file, black is 0, white is 65535, but 50% gray is 26982, not 32767. Other percentages of gray are likewise different from the value that would be found in the corresponding RGB version of the same shade of gray.

Why are the numerical values in monochrome TIFF files different from those of RGB TIFF files?