[NiHaoMike]

Work out what the power factor of a non PFC rectifier load would be if you gave it a square wave. (Answer: exactly 1)

Granted, that's a theoretical result only as a true square wave (with infinite slew rate) can not exist, nor is it even desirable. So make it a slew rate limited square wave ("trapezoid wave") where the slew rate is chosen to be reasonably fast, but slow enough to avoid EMI. The flat parts are effectively DC, but the transitions allow arcs to quench just like what happens with AC. Think of it as a compromise between DC and AC.

The question now, of course, is where would you find such a waveform? As it turns out, it's actually a very good waveform for a power inverter to output. During the flat parts of the wave, the transistors are simply left on or off as needed, with PWM only used to control the slew rate. That significantly improves the efficiency. It works even better with non PFC rectifier loads. The loads then only draw power on the flat parts of the waveform, requiring the inverter to only supply a small amount of current on the transitions to overcome capacitance.

It's also possible to design a generator to output trapezoidal waves, and in fact, the motor/generators in many hybrid and electric vehicles do just that.

Look into the future where alternative energy is the norm and rectifier loads like PoL (Point of Load) VFDs are everywhere.