It doesn't have to have 100% efficiency, it only needs to be a significant improvement over existing battery technology.
Example: my laptop battery has an energy storage of about 100 Watt-hours = 360 kJ. Suppose we want to store the same amount of power in an optical cavity (two mirrors pointed at each other) with a length of 0.5 meter. The time required by light to make a roundtrip in this cavity is 2L/c = 3 nanoseconds. Thus the power incident on the mirror at any given instant is (360 kJ)/(3 nanoseconds) = 10^14 watts!
This power is high enough that I suspect that even the vacuum itself couldn't conduct it without strange effects (production of particles, for instance).