The good news is the right-side rise seems to be gone. But I notice that both graphs have a bit lower EI and a bit higher contrast compared to XTOL and compared to what I was getting with earlier formulas. These differences are probably too small for folks to notice, but I'd like to know why this is happening. I'm mixing propylene glycol into the water to simulate a concentrate, and I'm wondering if PG reduces EI slightly, or if that's due to the very strong buffering, or if the citric acid could do this. So I'll try reducing metaborate and eliminating the citric acid, unless you have a better idea.
Again, the graphs show less difference than slight variations in temperature, agitation or other processing parameters will give you. If you want to lower pH without adding a new ion (citrate) to the mix, you could replace some of the sulfite with metabisulfite. Not sure it makes a big difference, but it should reduce complexity. If you reduce amount of metaborate, you also reduce buffer strength, i.e. you turn more knobs at once.
Trying to be the best of whatever I am, even if what I am is no good.