That's the exact opposite of what you should do. The more light pollution, the less read noise is a factor and the more full well matters, as the LP takes care of both swamping read noise and filling up your pixels with unwanted signal, meaning you should tend to favor lowering your gain. High gain trades off full well for lower read noise. When you start getting into these high gains with trivial read noise, you're just throwing away dynamic range for no good reason in the worst possible context. In any case, there's never a good reason to go higher than 120 on that camera in Bin 2 mode because it's not efficient in the trade-off. The only other gain to consider is the gain 0. Which you choose really requires a decision about what the minimum sub length you're willing to take is (e.g. perhaps you can't or don't want to process the number of subs that 120 would give you with the shorter subs it calls for, taking it out of the running). In Bin 1 mode, gain 0 is the only gain you should consider. It's fairly easy to determine potential candidates for gain: look at the dynamic range chart for the camera and look for the gains where it has maximum potential. Ignore all others. And never "just use unity gain". It's just a gain where 1 e- = 1 ADU. Unless your goal is to count electrons without a conversion factor, it's worthless as a guide to optimized subs. The only time you should be using unity gain is if it just so happens to be one of those gains where DR is potentially maximized.
LLM "AI", fwiw, is just a stochastic parrot. It doesn't know anything. I might broaden your search resources, as you did in coming here.
