You can use translucency on geometry with and without thickness. The results will be slightly different. Translucency is basically simulating very thin layer which scatters light.
With just one face on it's way, light will hit the surface, scatter inside of the very thin layer, and then pass to the other side.
With two faces, this will happen in two "layers", so the result will be probably a bit more "blurry", and less bright.
If you imagine a sharp shadow cast on such surface, with just one face, the shadow will be still sharp on the other side (imagine a sheet of paper). With two faces, it will get more blurry on the other side (imagine two sheets of paper, one placed behind each other).
Update: added a simple test to back up my theory. ;)