The reflected light is not a point light source, so the inverse square law cannot be applied.
Edit:
Since this is not so easy to see, I have a model at hand. Imagine an array of point light sources:
A real physicist should check, but I think that’s it. Quite embarrassing if not, so please be kind 