This business of "gamma", and the Photoshop color settings for Gray Working Spaces in particular has always confused me.
When processing RGB images in Photoshop, I can't see how the Gray Working Space makes any difference. I've done a lot of tests to try to find a difference. None found. So, please describe a test we can all do that demonstrates a difference.
Example 1: I open an RGB image and drop in some sample points. Then I cycle trough various Gray Working Spaces. There is no visible difference on screen, there is no change in the RGB values of the sample points. Same result in sRGB, Adobe98, and ProPhoto. Same result in 8bit and 16bit.
Example 2. I open an RGB image with the Gray Working Space set to Gray Gamma 2.2. I apply a strong Curves adjustment layer, flatten and save as Version 1. Then I repeat with Gray Working Space set to Dot Gain 20%, which is the default for my Photoshop. Same image, same Curves adjustment, flattened and saved as Version 2.
Then I load Version 1 and Version 2 as layers, put the top layer in Difference mode, and check the Histogram StdDev value. It's zero. No difference. Again, the same "no difference" result regardless of RGB working space and bit depth. I've repeated this test with different types of adjustment layers, never finding a difference.
Example 3. I open an RGB image, drop some sample points, and dupe it. With Gray Working Space set to Gray Gamma 2.2, I convert one copy to grayscale via Image/Mode/Grayscale. The I change the Gray Working Space to Dot Gain 20% and convert the second copy to grayscale.
The images look identical on screen, but the sample points (K% values) are now quite different. Why is that? How come different sample point values produce the same on-screen values. I convert both these images back to RGB mode, and Voila!, the sample points are the same again.