The square footage is relevant when estimating water used and that is it.
This is tricky and a sticking point with a lot of ppl, some pretend it is easier than it is.
The tricky part is, and I'll do a horable job articulating it, nozzles mean nothing. I can take the same nozzle that flows at 5gpm and provide a nozzle that will increase the pr but shrink the coverage or decrease the pr and increase the coverage. So now you are thinking, doesn't matter, 5gpm X 4 nozzles is 20 gpm using area we can get a pr, so who cares, and I'll agree for the sake of argument. The issue arises when you take into account head pattern. You can follow a square pattern or a triangle pattern, with the same number of heads this will change the pr of the yard without changing the quantity of water applied or the number of heads, so a flex daily schedule will perform differently for all of the same values except nozzle placement.
You could account for the pr difference by adjusting the efficiency of the zone which indicates how uniform the water distribution is, thus flex can over apply water to ensure that lower pr areas reach their saturation. But now you have provided a setting that is a fuzzy setting, you are still stuck with using catch cans to compute the pr of multiple spots in the yard to provide an accurate value for efficiency, so you just dragged an accurate pr measurement back into reality, thus escaped nothing.
Now let's consider the tolerance to the system. if you are looking for accuracy, you will need to measure the psi at the heads and plug that value in as that will have an influential impact on the actual flow rate of the head.
So I will agree that nozzle count with gpm will work on a wasteful schedule like a fixed, and could be implanted for that specific but will be very counterproductive for flex.
Now here comes a curve ball, in 1 spot of my yard I have a 3gpm head with 4 5 gpm heads. I have to do this to keep my efficiency high, so technically this solution would lie in this setup by 2gpm of every active schedule minute which would be 30 gallons a cycle over 7 cycles in August on a single zone. Maybe inconciquential, just another data point.