I'd say this is walking on thin ice already,
Not a good place to generate a lot of heat :-)

Erik: I'm not talking about running 10A through a device that provides 1Kohm resistance at that moment (that would be about 1000V) but simply over short time (100ms?) decrease the resistance.<>/i>
for simplicity let's assume a resistive load. !A @220V = 22 ohms.

So at the midpoint you have the limiter @ 22 ohms = total 44 ohms = 5A and each take 110V, so it is 550Watts. Even for a very short time, that would fry most devices.

Anyhow, if you try it and it works, do not believe that's it. You can easily get in a situation where the device weakens just a bit every time it happens and after 1000 instances finally break.

"succesful testing does not prove the abscence of bugs, it proves the abscence of known bugs"

Erik