If I understand correctly, it's not the light reflected off the exoplanet that we detect, it's the slight dimming of the stars light as the exoplanet transitions between the star and our telescope.
I'm not sure what this has to do with a communications link though. It's not like we need as much energy as a star, we can use a directional link like a laser. Sure it'd still need to be massive but while I don't have the math in front of me, it seems like something we could build.