Blowing the dust off the thread... Class C is more efficient and is appropriate for High Power design if linearity isn't your first concern. [see N. Dye, H. Granberg, Radio Frequency Transistors, 2nd Ed].
Mack's test proves a good point; at a given amperage with a specific power in the AB bias will produce more watts.
What's being lost in your efficiency argument Booty is, for a fixed amount of amperage (i.e. you have one alternator that produces 200 Amps, or a single power supply that produces 100 Amps) the efficiency of Class C will give you more watt's ultimately than a class AB.
If we wanted to accurately measure Macks test and compare apples to apples we'd also need to measure the amount of amps drawn at the time of the test. What we should see is the Class C amp requiring fewer amps to produce nearly identicle watts out - leaving more capacity that will eventually surpass total power out that can be produced by the AB platform.
If efficiency was of no concern; than amp builders would all use Class A and AB and AB1 would be the red-headed step children of RF design...
