I wasn't particularly worried about the operating class of the amplifier. The datasheet for the MRF-101 makes it look very linear at 13.56MHz with IDQ of 100mA up to 100W:
My circuit adds negative feedback that further linearizes the amp. With IMD products between -38 dB and -41dB I'd say that it's better than you can get with a Class AB. Maybe if the bias was increased, the linearity would improve even further but I haven't made any measurements at increased bias levels.
Jim.