I had an idea this evening and wanted to run it by you lot.
AC input of 0-100v under no load, 0-6v under significant load. Looking at the pic, the only theoretical disadvantage over a full wave skhottky rectifier is the darlington will drop 0.9v instead of 0.55v. The benefit is when no load there's no heat to shunt (to protect the buck circuit that would be connected to +/-).
The big issue I foresee is the Vceo which is rarely found above 25v and typically 7v. I can't see anyway around it so another schkotty for reverse protection on the emitter would drop a further 0.55v. Ultimately then the voltage drop would be 1v greater than a zener shunt -> full wave schkotty -> load when a load is connected. This would be too great an efficiency loss over the no load benefit.
Any improvements or thoughts?