I don't think I said anything about simple square or modified sine inverters, as I said the Victron has a PWM-controlled FET bridge right off the battery.
I think there is a terminology issue, you define LF as switching at AC frequency i.e. 60 Hz or some low multiple of that, and HF as anything that uses PWM. I'm not sure that's a widely accepted definition. Is there a standard definition here? If there's a reference I'd appreciate it.
I'll have to find the reference I'm thinking of but I thought there was a distinction between PWM switching large currents at low voltage direct from the battery (i.e. as Victron does), vs. first boosting to high voltage DC and then PWM switching there. The first has much larger FETs due to switching such high current, and as a result a lower switching frequency to keep switching losses down. Switching frequency in the 10 kHz range. The second switches high voltage at low current thus smaller FETs, smaller magnetics, higher switching frequency in the 100s of kHz to MHz range. That's what I meant by LF vs HF, but I certainly good be wrong there.
I don't know what all the limiting factors are on surge capability in these designs but I'm guessing thermal losses in the transformer core could be some of it, and if so then a large mass transformer will take longer to heat up and thus have longer duration longer surge than a HF inverter with small low mass magnetics.
I'll look into this some more, I am curious. I'd appreciate any references on this topology difference with respect to surge capability.