I did not consider that, thanks for pointing it out. Here's the order of the test. The AC hadn't been running until I started the testing:
In one session I did:
Fan On @ 400 CFM/ton
80% of 400 CFM/ton
This took ~20 minutes total. The time to note the data and swap the filters was just a couple minutes. Then I shut everything off, reset the switches, turned it back on, waited the 5 minute tstat delay, and got back to it.
80% of 350 CFM/ton
15 minutes for this step.
I did the Fan On @ 350 CFM/ton at a later date.
So that effect perhaps resulted in greater static pressure as I went through the steps, but it does appear that the effects of CFM and filter type dominate. That effect would also support the ESP-mitigation steps I outlined.