So I finally found out what my invention is called. Seems like back in the 80's someone had the same thought as me and called it the Kahn-Hazeltine system. Not a polar AM yet it serves a common function.
So this is the final version of the modulator. I've changed the oscillator since the one in my previous circuit was working only because of an EC glitch and wouldn't have worked in reality. I also found out that it has a few disadvantages over other stereo AM techniques. Looks like it has a very poor stereo separation at >5Khz reaching to none at 7Khz. Still it has a lot of advantages over other techniques like simplicity better use of the occupied channel etc.
Just added the modified detector circuit you'll need to demodulate and restore the original signal. Since the transmitter is modified so is the receiver.
