That is correct. Not 10-22KHz as you orginally wrote. Models vary but mine has a 22-22KHz filter: you don't have to use it.

It looks lke the nulling oscillator runs out of gas above 100KHz.

I miswrote earlier. My SG502 measures measures 0.0190% at 100KHz, not 0.190%. Measures 0.0078% with the 22KHz filter in ;-).

EJP