actually, i find the FTC requirements arbitrary. Music is not a 100% duty cycle into an 8 ohm resistive load. Having short term current and power capability is far more important than being able to, say drive a 75 watt signal into an 8 ohm resistor for an hour.
Well recorded music has a peak:average ratio of over ten. This means that, to avoid clipping, an amplifier ought to put out less than 10% of its maximum voltage swing on average. OTOH it is VERY useful for it to have the ability to drive nearly zero load momentarily, given the very reactive properties of most cone speakers (with the voice coil flying in the reverse direction its theoretically possible to be negative in fact).
The simplistic requirement is a bane, IMO. I realize it is nearly impossible to make anything simple meaningful, but this really does not bother me one bit. Not one’s allowing them to lie, simply to stop measuring "20-20k, both channels driven, for an hour, into a 8 ohm resistive load". Of course, I’m one of those misleading designers intent on selling you something that sounds good, but is really a pile of crap :-)
Note the biggest hurdle is the most fallacious: its heat dissipation.
note2: for any class-A amp this is all irrelevant and for hgih bias AB far less relevant.
note3: read specs at 8, 4 and 2 ohms to get a good idea of real capability.