Tom, for the sake of argument, let's take a step backwards from all this talk of photons hitting metal, and the mechanics of RF antennas.
The point is, it's been demonstrated time and time again, in the real world that it works. By millions of people. When the location of a transmitter is known, let's say, a television transmitter, it's well known that adding a rooftop antenna if you're in a fringe reception area along with a motor for rotating the antenna, will allow you to improve your reception by "aiming" the antenna in the direction of the transmitter. I have experienced this myself, while attempting to receive broadcasts of football games that used to be 'blacked out' locally, but I could often receive them from distant cities not affected by blackout rules by aiming my directional antenna towards a distant transmitter. The difference in signal was more than just a little noticable, it was the difference between a viewable braodcast, or none at all.
If, as you say, these types of antenna are not in the least directional, what then accounts for this observed behavior?