newbe - amplifier distortion, a bit long...

I'm trying to get a handle on amplifier distortion. It seems to me that the amount of harmonic distortion is not as important as the order. Low order being acceptable and high order being objectionable. So long as the total is relatively low say under 1 or 2%. I make this assertion after measuring and listening to several different amplifiers. Also, in looking at the measurements in Stereophile magazine, I find no correlation between the measured performance of the amplifiers and the reviewers subjective impressions. The only correlation I find is the more expensive amplifier always sounds better! uh yeah, right.
So, if it is true, I wonder if there was a perfect amplifier with NO harmonic distortion at all would it sound good? Some people might say “well the amplifier is perfect so you can’t blame it”. But I say if the perfect NO distortion amp makes everything sound like crap what good is it?
And so my dilemma, when designing my amplifier what kind and how much distortion should it have or does it matter at all. Either way it seems to me the generally accepted practice of designing amplifiers for lowest practical distortion does not have any proven merit.
If you think I am a lunatic, please kindly, tell me why I am misguided.
 
sam9

I would think you need to first decide what you want an amplifier to do. There is the tradition "straight wire with gain" ethic which pushes for the lowest possible distortion in all categories, IM, THD etc. Then there is the desire for an amplifier to have some sort of character, "musical", "warm" , whatever.

There are also functional constraints. Size and weight are among the more mundane. Power is also a factor. If you have super efficient horn speakers, a 200W+ amp is kind of wasted. If you have a some huge inefficient speaker like really big Magnepans you won't just be looking for lots of watts, but lots of watts at 4-ohms.

So the question is: What do YOU want?

If you are clear in your want/need then your questions may be refined enough to already contain the answer.

Oh, yes. Don't believe everything you read - especially when you already agree with it!