What should I be measuring it to? between the ground plane and the ground pin on the chip?

I have a few of these: https://www.digikey.com/en/products/detail/epcos-tdk-electronics/B64290L0044X049/3913505
Would these be good for common mode filtering?

I currently just have the primary wires running through a random small green ferrite ring a single time. do I need more turns?

In particular look at ground pin at chip.  Look for very short spikes or ring at high-to-low transitions of output.  Good to check power pin too, but that is less likely to be a problem.

The most useful place for a common-mode choke (ferrite ring) may be on the high-side output.  That winding experiences by far the most rapid and highest voltage swing.  If you have another ferrite ring, I'd add that to high-side output, ideally with 3 or 4 turns of the twisted-pair secondary output leads.

Coil was running at ~400kHz, with very low duty cycle (a few milliseconds and like 2 BPS for testing). no noticeable heating in the gate drivers before they died.

When my new chips arrive I will try scoping the pins during operation to see if theres any weird voltage spikes.

I added a ferrite ring to the input of the GDT to hopefully block interference, but I'm out of 27524 chips so I can't test if that helped yet.

I was gonna use compressed air through the spark gap column to extinguish the arcs. Is that not necessary if I have a switch controlling the 24V from the batteries connecting and disconnecting it (24V batteries > ZVS driver > flyback > marx gen)?