OOPS... I thought I answered it but somehow the post didn't came up...
It should be slightly more than 1 stop of light loss, if the incident light is circularly polarized(isotropic). (Ideally it is 1 stop but there's always extra losses here and there. )
However, light reflected off non-metalic surfaces are ovally polarized, meaning the intensity of EM wave is different in different directions. And how 'oval' the light is depends on the incident angle of that reflected light beam - and the ovality and direction of your PL front plate determines how many stops are lost.
Drudkh - good point it's tested to be slightly over 1 stop with a Hoya 67mm CPL, against a blank wall.