Hi Eleanor and Cliff, I've got a bit of a dud connection today but I think I've read and skimmed a fair amount of yours and Cliff's site and docs.
Re the "modulated" microwave. I think you are saying that the early claimed successful demonstrations used a pulse of MW. Presumably some trials were done to determine the duration and the amplitude of the pulse and that particular parameters were chosen so as to allow a reasonably safe but effective trial. And that the envelope of the pulse when represented graphically and schematically is drawn as a square wave. The "modulation" is done by way of timing pulses to coincide with the instantaneous frequency of the sound wave. So in the ordinary way, a sound wave is transduced to it's electrical analog. The zero point is established and the MW pulse is fired each time the electrical signal equals zero.
So we might assume that a person would sense something if, for example, the pulses were delivered at 1000 hz. And if so then at 2000 hz the person might say, "Higher". If this is true ie that a person can hear anything at all and that they can then distinguish tones it is plausible that "modulating" by way of timing a uniform pulse to coincide with the audio wave could actually convey information to a person. If a person was totally oblivious to having been targeted and then experienced something that was like but not quite like speech and they could hear words they would be startled to say the least.
Have I pretty much got it?