Many of the responses here are spot on. Ralph's are definitely spot on. And in no way was he touting his products over others. Amp design is formula based. Anyone that has taken amplifier design in college knows this. Design an amp that is nominal 100 wpc, drives typical 8 ohm loads, has input impedance of x and output impedance of Y, input sensitivity of z, operates at a frequency range from DC to light, etc. Pick the devices you want to use. Tube, solid stated, hybrid, BJT, FET, etc. Output impedance is a formula. Input impedance is a formula. Gain is a formula, frequency response is a formula, etc.
Many people take other's designs and use them to build amps. They never did the calculations and detailed design work and therefore are unaware of those aspects of the actual design.
A well designed amp is just that, a well designed amp. I'm not quite on board yet with Class D designed amps. But, I'm open to great Engineers/designers and maybe one day I'll demo one that is really nice.
But, Class A, A/B, etc. Tube, solid state, this all depends on how good the designer, circuit topology, design, construction, and parts are.. My view is that even class A/B amps can really sing. Listen to a Mark Levinson 23.5 that is to spec and tell me that isn't one of the best amps that came out of that company.
The power supplies have to be well done.
If one has inefficient speakers, then lower power amps would have a really hard time driving them well. With efficient speakers and a decent room, 5 to 10 watts would almost drive you out of the room. So a good low power class A amp would be great. If I had more efficient speakers, I would grab a pair of Mark Levinson 20.6 amps in a minute.
As mentioned previously, most class A/B amps are actually class A to a set power rating then convert to A/B status. So, a lot of people still confuse class A status. if the amp's output stages are always turned on, then that is class A period.
The costs difference is typically based on a few things. 1. Power supply transformers, number of output drivers and lastly, heat sinks.
Take a Mark Levinson 23.5 amp. With the power supply internal to that amp, two massive 1500 kVA (I believe) transformers, then only reason this isn't biased as full-on class A, is the costs of the necessary heat sinks. Which would be stupidly huge for full biased 250 wpc.
Nelson Pass has an older write up/paper on power supply design and specifically discussed transformer and heat sink requirements. I believe it is on his DIY page. lots of really good reading there.
but, does Class A amps sound better? Yes and no. depends on the design, speakers used, room, wiring, etc.
Apples to apples comparison. Say your speakers sing with 100 wpc amps as recommended by the manufacturer. Say you pick an A/B amp and a Class A (fully biased) amp that can specifically drive your speakers. both solid state or both tubed. Both are within the same price point (important) and are well designed amps. I would bet you would have a really hard time hearing the difference at low power levels, because they both are probably operating Class A at that level anyway and at higher levels, you might, just might hear a slight difference. But I doubt it.
Sorry for the long post, but this is fun.
enjoy