As a singer and part-time engineer who's recorded a fair bit of folk/acoustic oriented music over the years, I'm of the opinion that if you're aiming to create a natural-sounding recording- ie a representation of a "real" performance- you shouldn't need to resort to such trickery. But anyone who's spent time in a studio will know that recordings are often made in less-than-ideal conditions, under time pressure, by human beings who have off-days. Pitch correction is one way to get more consistently good results in those conditions and properly done, it can be pretty transparent, as well as a darn sight quicker than re-recording the same line time after time. Some may see that as dishonest, but is a single live take with a bit of pitch correction here and there somehow less honest than a "comp" made out of bits of seventeen different takes recorded verse by verse, line by line, syllable by syllable? Both create the same illusion- that of a single, "perfect" performance. I'd venture to say it happens more than you might think, even (shock! horror!) in folk music.

At the other extreme there's the deliberate use of Autotune as an effect in its own right, a different matter altogether. Any technology opens up new creative possibilities. The "robotic voice" effect is a bit old-hat now (Cher's I Believe was 12 years ago) but there are loads of other less obvious things you can do with it- it's not just for vocals.. The real problem is the routine, unquestioned overuse of it, regardless of whether it's actually necessary. Producers these days are becoming like the proverbial fool with a hammer who thinks everything's a nail. Because you *can* produce perfectly "in-tune" vocals, people think they *should*. Then everyone gets used to the sound, and forgets what a really good singer actually sounds like. And someone who can't hold a tune in a bucket finds they can sound *almost* as good with a box of tricks, and no-one trusts a perfectly good singer to actually be in control of their own voice.