A script is 100% accurate, it reproduces the same movement every time. I also have an emulated USB mouse that does the same. If the sensitivity changes, the script will obviously show it. It's impossible for it to reproduce the same result if the sensitivity is changed. That would make no sense.
By hand you are hard pressed to detect an obvious difference of less than 1%, I'm not exactly sure what you mean here? 56 for targeting can be for instance 72.0037 cm 360 distance, and with the same hipfire sensitivity this means that 55.456 is 72.71 cm 360 distance. I don't think a 7 mm difference is very obvious to test by hand, but it's actually pretty easy with scripts.
Changing the GamepadLookAcceleration values has absolutely no impact on mouse sensitivity.
Smoothing and acceleration is already disabled in this game, adding those lines does nothing.
Actually, changing anything in the file is pointless, the game doesn't read from it. If you want an obvious test, set the hipfire sensitivity to 0.1 in-game. It's really slow. Exit the game, change the MouseLookSensitivity to 10, save the file and set it to read only. Start the game, is the sensitivity really fast now? No, not here at least