Yeah I understand how a GFCI circuit works...
There are different quality of GFCI devices, different GFCI chips and even the circuits can be single coil vs dual sensing coil depending on application. As we know all it takes is 4 to 6 ma (5 ma) delta for over 2 ms by typical design, this can be caused by
any circuit of the downstream devices,
including crappy low quality GFCIs. All are induction based. YMMV.
The common Fairchild chipset data sheet:
BTW that chip can use a single coil design by looping the test circuit back through the sensing coil instead of bypassing it, then the second coil is "not needed".
Another potential false trip item, if you look at the circuit the drive to the GFCI solenoid is via an SCR, SCRs of course latch once triggered and we want it to here! (Latched until all power is removed). This can also be a cause of problems if there is noise on the electrical line as it can trigger the SCR to trip (false positive), in the circuit there is a noise shunt in that part of the circuit to prevent this in theory but real life can be different based on quality and/or C3 can wrong sized or be failing. It is another one of those can't figure out why GFCI keeps tripping here but if I swap it to another line it does not fail.
Besides the fine electronics, as pointed out it is wasteful and beyond even that, test one and then find out which one(s) tripped...