Doug at Challies.com announces his preference for the ESV over other Bible translations. I'm not going to comment on most of what he says, but I found one thing very interesting. He gives a number of different translations of Psalm 119:11. You can check the post for the different versions, but here's what fascinated me. Aside from The Message, the most dynamic translation out there (done by a real scholar or scholars, anyway), the two most dynamic translations of that verse were the NASB and the ESV, which are normally much more formal translations. All the others retain some word for hiding when they record the psalmist's statement that he's hidden God's word in his heart. The NASB and the ESV use other words. The NASB psalmist has treasured God's word in his heart, and the ESV psalmist has stored it up in his heart. I just found that interesting, but it underlies an important point.
The NASB tends to be the most formally equivalent of English translations, to the point of not sounding like English. The ESV does much better at sounding like English but isn't quite as formally equivalent. Yet on individual verses, other translations will turn out to be more formally equivalent, because every translation will make decisions about when to be more dynamic in how to translate particular expressions. In this case, the translators decided it would be misleading to talk about hiding God's word, as if that means they're keeping it so others can't find it. That's why dynamic translations do not alter the word of God. They make clearer what the word of God says.
Update: I found Rodney Decker's review of the ESV, which points out a number of places where the ESV is probably accurate to reflect the meaning but has supplied words where the original does not have them, and it does limit the exegetical options. This sort of illustrates the same point. Even the more so-called literal translations will do this, so it's best to have multiple translations for careful study. Yet this also shows that the functional equivalence or dynamic translations are just doing a little more of what the formal equivalence translations do all the time in ways that people who call for a literal translation will not like.
Decker has some very effective arguments against the kinds of claims you'll find in such literalists as Leland Ryken, whom Decker claims just doesn't understand Greek well enough to know that the best way to translate a subjective genitive into English will require what Ryken would call supplying words not in the original. Of course, every word in the English is not in the original. The question is how many words are necessary to translate a subjective genitive, which is only one word in the original. The truth is that it will be more than one, usually a few.