It doesn't matter. You have square root of 2 in the numerator for the voltage and a square root of two in the denominator for the current. They cancel out. You can use whatever as long as the ratio of voltage to current stays the same.
------------------------------------------------------------------------------------------
If you can't explain it to a six year old, you don't understand it yourself.
As HamburgerHelper said, rms is OK as long as both V & I are the same. Units do matter. If f is in Hertz, V is volts, and I is amperes, then L will be in Henries
Keep in mind that the equation is only correct if there is no resistance in the circuit.
I want to use it to get a rough feel of the magnitude of the Inductance for a transformer. If the magnetising current is 5A and the primary side voltage is 33kV.For a 50Hz system.
Here are some common R% X% Z% values for transformers of different sizes. As you move up in size, X% makes up a larger portion of Z%. Enough so that for a rough ballpark number, you would not be off by much just to assume Z=X.
This doesn't matter unless you are doing high frequency transient analysis but there is capacitive coupling between the winding and the phases. The impedance of a transformer is a function of the frequency. High frequency transients can pass through the transformer through capacitive coupling without stepping up or down through winding couplings.
------------------------------------------------------------------------------------------
If you can't explain it to a six year old, you don't understand it yourself.