As I mentioned here, translations can indeed be both better and worse than the original. Which, if you like your probabilities and mathematics, is a statistically guaranteed outcome.
Of course, quality is subjective, but if 'enjoyment' and 'popularity' are anything to go by, then you can start to see obvious examples of the same-if-not-better translations emerge.
And who can say how good the original was if one cannot read it? Which reader of the source language can truly tell how good translations are? Those able to do so are few and far between, but other, more accessible metrics can at least show us how 'not bad' translations and originals are.
Funny that literary translation is still so poorly rewarded financially. Shouldn't there be a publisher who specialises in highlighting and fairly remunerating these unsung authors? Perhaps the new e-publishing wave will enable this with its lower cost barriers for promotion.