Just this little sidenote, or big sidenote:
Those early phasors were "crude" in the sense that they were simple by today's computerized models,
but an engineer from who's knowledge and experience I learned much, shared with me that
when an array is aligned as much as possible by adjusting the lengths of transmission lines,
rather than by the quick and easy method of big coils and caps, it will be much flatter across the full channel.
Use the coils and caps for fine tuning and do it as much as possible at the LTU shacks,
not the transmitter room, and the entire system will be less reactive along the transmission lines.
Big coils and caps exhibit much higher Q's and narrower bandwidths than long lengths of cables.
I participated in an antenna proof of performance and was very impressed at how undistorted the station was,
even in its deepest nulls, and they were and are deep, it was completely undistorted.
In other words, the sidebands across the channel were just as suppressed as the carrier
and the audio was good in all directions.
I often have wondered whether you could just have separate linked transmitters locked to a single oscillator frequency with computer generated power level/amplitudes and phase shifts driving each tower independently, rather than a phasor. No doubt, some tweaking would be necessary just like Barry Mishkind and Ron Rackley have described with WFLA WSUN with the mutual impedances, but I'd like to see this tried if it hasn't been done already.