Mark - When you listen to 200W peaks your class AB amp really delivers to speakers on average only few percent of peak power. It is because music is not always at the peak and half as loud means 1/10 of power. In addition there are gaps (unless you listen to sinewaves). Nobody sane would design power supply in 2x200W amp for 800W required (50% efficiency at best). Typical amp can deliver full power at limited period of time (like 1 minute) before thermal protection kicks in. There is no need to design large heatsinks for the same reason - otherwise they would have to be the same size as in class A amps.
Class A is extremely wasteful but has one big advantage - is more linear before negative feedback is applied. At the end (with feedback) both class A and class AB might have the same linearity but class AB required 20x more gain (before feedback) to achieve this. This huge gain (in order of 4000) in connection with limited response of the amp (limited slew rate) is causing late feedback correction. In time domain it shows as small overshoots of pulses and in frequency domain it shows as enhanced higher order odd harmonics (bright sound). This is called TIM (Transient Intermodulation Distortion) and was unknown until 70's. There are great class AB amps where designer minimized the problem and also some bad class A amplifiers. You have to listen. My general intuition is that better specifications mean worst sound (achieved by deeper feedback). Same with power - what is on the paper is useless. The only usefull parameter you can find there is weight (some correlation with power).
Class A is extremely wasteful but has one big advantage - is more linear before negative feedback is applied. At the end (with feedback) both class A and class AB might have the same linearity but class AB required 20x more gain (before feedback) to achieve this. This huge gain (in order of 4000) in connection with limited response of the amp (limited slew rate) is causing late feedback correction. In time domain it shows as small overshoots of pulses and in frequency domain it shows as enhanced higher order odd harmonics (bright sound). This is called TIM (Transient Intermodulation Distortion) and was unknown until 70's. There are great class AB amps where designer minimized the problem and also some bad class A amplifiers. You have to listen. My general intuition is that better specifications mean worst sound (achieved by deeper feedback). Same with power - what is on the paper is useless. The only usefull parameter you can find there is weight (some correlation with power).