I assume this terminal fishery spreadsheet model works like nearly all others. The fishing season harvest rates get applied to the same terminal run size forecast. There is too much overlap in time and space for the model to act like gauntlet fisheries where fish caught early and down low are subtracted from the starting run size for estimating the number that can be subsequently caught later. A test of this in your spreadsheet is if the QIN harvest changes when you manipulate the NT fisheries. If the QIN harvest doesn't change then the spreadsheet model is set up the conventional way which is easier to work with.

I looked at the model summaries posted on WDFW website. They are laid out with run sizes, harvestable shares and estimate catches per fisheries. Assuming the spreadsheet model is conventional, if you simply shape the NT fisheries to produce the harvestable NT share you get an idea of the NT fishery options. Ignore the escapements, just work the NT fisheries to abide by the NT fishery rules in the GHMP and get it to yield the NT harvest share.

Yes, freeing up the QIN harvest piece would allow for getting all the numbers to be green. But messing with QIN fisheries by NT fishers would be objectionable to QIN, as it would if QIN shaped the NT fisheries. There's already too much of "we don't like what you are doing with your fisheries, you need to do this". Having a model that allows for both sides to manipulate the other just encourages this.