I work for the NRAO but I am not a radio astronomer, but my sense of it jibes with yours. Planets are mostly so close to their host star from our perspective that to distinguish signal between the two would require extraordinary sensitivity. But that sensitivity also means that you have to point very precisely. A less precise instrument can cover more sky at once but then you can't pick out individual sources as well and brighter sources (like stars) would swamp weaker sources like terrestrial radio/television signals.

I don't think we have enough information to tune the values in the Drake equation to make meaningful predictions. We'd need observations to do that, but the observations are, as you point out, difficult enough that we don't know if they're absent or we just can't see them yet.