Oh, you're not very charitable in the conversation. That's too bad. I'm not a conflictual person, I won't engage in a rhetoric battle to look wittier.
Do you actually believe your answer adresses everything I said?
You can affirm that "the purpose of audio reproduction is exactly that: faithful sound reproduction" but it's too easy to affirm that is the end all and be all of the discussion.
Let's assume that it is. It doesn't forbid to question it.
If I ask "why is it its purpose?" what will you say? "so that we can have the truest experience of music?" and if I ask why it matters to have the "truest experience of music", what will you say?
(Any argument based on "how the artist intended it" is inherently flawed, I hope you know that)
If Your question really is "Why should the ASR insistence of faithful reproduction of original recording be the desired goal?", here is my answer:
First of all, it does not always have to be. For example: If the consumer's goal is "self-satisfaction achieved from shopping", the ASR point of view is probably not critical, but I think it should be obvious that these kind of cases are not meant to be within ASR scope anyway.
Concentrating on cases where the actual, or perceived, sound quality is what matters, I'd say we can divide it into two cases:
a) The consumer actually wants the faithful reproduction of the original recording. In this case, I think the merit of ASR approach is self-evident.
b) The consumer wants sound that pleases him, never mind the faithful reproduction. I will argue that the ASR approach is still the best option for him, because it is much easier to start with a faithfully reproducing system, and then use signal processing to modify the sound to Your liking, than to start with something else.
With a faithfully reproducing system, You have a known foundation that includes all the information needed. If Your system has already lost part of the information (via limited frequency response bandwidth for example), there is in practice very little You can do about it - even if satisfying You personal taste would require it.
As for the other basic parameters highlighted in ASR - distortion and noise - the thing to understand is that it is much easier to add them afterwards, if You wish to have them, than to take them away once they are induced in the signal. A well done steak can not be changed back to medium, and a brain tumor is extremely difficult to remove without removing also something You definitely wished to stay as it was.