But what about bigger capacitors? 
And higher voltages?
One thing to consider is that beyond the capacitance and voltage, when I transient test the meters with the low voltage generator, the edge rates are very slow. I suspect the grill starter damages so many UNI-T meters in part because of the edge rate. To get the UT61E to survive, I gave up some of the meter's AC bandwidth. When you jam a meter across a capacitor like this, you may create a fairly fast edge.
As joeqsmith has figured out, almost any meter that survives even moderate transient tests should be fine.
Of the meters that survived the 5.9kV 100us FWHH and 2ohm source, the only meters I did not check against the capacitor bank were the Fluke 107, Brymen BM786, BM839, Yokogowa TY720 and UNI-T UT15C.
The UNI-T UT15C was damaged beyond repair and recycled.
The BM786s are early prototypes. I had done a fair amount of rework on them. I am very confident they would survive as they use the same front end as the 789.
While the Yokogowa had been damaged, the front end was fine (nightmare to repair). This meter had no problem discharging the 1uF 2kV capacitor bank.
The Fluke 107 was damaged at 15kV. It damaged one of the high speed clamps and looked like it had caused some damage to the PTC. I did repair this meter (eventually replacing the PTC). It had no problem discharging the capacitor bank.
So, I'm going to go out on a limb and stick with my original statement that if a meter makes it though all of my testing including the low voltage transient generator, it will survive this new test.
I was a bit surprised that the Brymen BM27s (rebranded AMPROBE PM55A) has held up so well to everything I have put that pocket meter through. Remember, member True had four of these that all failed. Mine arced over at 8kV but no damage was done to the electronics. Impressive for a little pocket meter.
https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg1787117/#msg1787117