Jupiter recommends usability over personalization. Because personalization is more costly, and they haven't been able to capture the benefits quantifiably.
This is a pretty shallow argument from Jupiter. It's kind of like the whole IT-productivity debate. It's not that there's no relationship, it's that the question itself is asinine. There are so many intangibles and feedbacks among the variables that it's irrelevant.
It's also a matter of strategic choice: do you want to be Google, or do you want to be Amazon? Either strategy is valid - but being stuck in the middle is where the danger lies.