No, really, why are they? You know, I don't mind it when critics point out some flaws in overall good movies, books or pieces of music, but when they begin to complain about the tiniest details, I'm not happy. And unfortunately, that's what most critics do. I can think of some exceptions (Steven Greydanus of the National Catholic Register, T.S. Eliot when he actually wrote literary criticism and a few others) but the vast majority seem too concerned with basing their criticism on criteria of originality, or structure, or whatever else they might choose. When they start to criticize movies, books, and pieces of music on picayune details that aren't even objectively important, critics just look silly. Or to some, they might look intimidating, because many people nowadays think that if people use big words and are picky about details that no one else would notice, they must be extraordinarily smart.
This criticism of critics' criticizing style aside, however, the question of what makes a work of art good is undoubatedly very debateable. As a friend of mine said, "If you pursue to its logical endpoint the idea that one piece of art can't be better than another, all art is created equal and there is no way to determine the good from the bad, because there is no good or bad." Yes, that's a definite conundrum. A complete relativism on the subject is what causes people to call rap, "poetry"; to say that corruptions of language are natural and should be welcomed; or so forth. On the other hand, an overly critical attitude causes reviews like the one I just read on Crime and Punishment - it claimed that the book was worthless because it wasn't structurally flawless.
It seems to me that there must be definite standards out there. It's just defining and agreeing on these that is the problem. Often this is where criticism just collapses into opinion. For example, if you're going to say that structure is what defines the quality of a book, then you can certainly condemn a classic like Crime and Punishment - the only thing is, practically no one will agree with you unless they happen to agree with your criteria.
CS Lewis' An Experiment in Criticism is very unlike any other literary criticism theory I've heard of, but I think it's quite plausible. He had the idea that a work of art is defined by the people who read (or listen to, or admire) it. So, those books most admired by "true readers" (those who don't read egotistically - simply in search of thrills or of "egotisitical castle building", as he calls it) are books which are good.
A full time critic of any form of art is sure to be biased, no matter how much he tries not to be. If this bias coincides with the absolute standards we can't seem to pin down, then this isn't necessarily a bad thing. But most critics will be looking for a specific thing in art that they personally consider essential to it. They also tend to be heavily influenced be intellectual fads.
Now most critics seem to be obsessed with originality in art. If a book is "different" it is defined a classic. A few decades ago (at least in the forwards of my older books) critics seem to have been preoccupied by structure. But the literary public often as not disagrees with their assessment, because the criteria rings true for only a few individuals. Half the population looks at cavils against a really good book and says "who cares!"
If the book is well-writtten, Lewis thinks, the literary public (not the romance or horror novel public) will recognize it as such. And I jolly well agree with him. It sounds like the most sensible criteria, even if we haven't yet agreed on a definition of what properties of the art itself make it good.