It might lower the power consumption partially because it is sending to control signal, but the servo is still powered.
I guess the terms "sleep a servo motor" should be more technically defined. There have been other previous discussions of removing total power from a servo. when a servo is detached there should not be any control signal going to the servo control chip, which would result in the servo h-bridge not receiving any control pulses, resulting in the servo motor not receiving any power, which could be equated to the motor being "asleep". The servo chip and the pot in the servo are using some power anytime the servo itself is receiving power.