No, because of quantization. (the pixel values only have discrete values, not infinitely precise values)
And the middle slider is a gamma factor. x = y^gamma or x = y^(1/gamma) depending on how you look at it.
Thanks. I don't really understand that but it sounds like there would be less image distortion doing this if in a 12 bit image.
Jim, just remember that gamma(x) followed by gamma(y) is equivalent to gamma(x times y).
To reverse a gamma(x) adjustment, apply gamma(1 divided by x).
Work in 16-bit mode for greater precision than 8-bit mode. (Infinite precision would be required for perfect reversal of gamma adjustments but 16 bits will be good enough for most purposes.)
And with multiple stacked Adjustment Layers even just changing an 8-bit image to 16-bit (without having access to the raw data) can alleviate some quantization effects somewhat.