I vaguely remember an article from around the time RFID passports were a hot issue, in which researchers used multiple capturing devices and were able to square the reading distance. I don't know whether that was specific to the distance they used, the type of RFID, or even an upper limit, but it was an unbelievable improvement.