I believe they exist, sure.
Like most forms of evaluating people, they're going to catch some things that other observations miss, and miss some things that others catch. Good analytics will try to figure what it got wrong and why, but everything has its blind spot. Smart organizations use many evaluative tools, and some times these tools will disagree about a player. The key is in reconciling those differences. In the old days it used to be two scouts having opposite opinions. Now it might be the scout vs. the computer, or even two different predictive models. If you're drafting a player, or signing one or trading one, it's a binary decision. Either you get him or you don't. Unfortunately the data is always probabilistic, so you're going to get it wrong sometimes.
Analytics is used to get things wrong less. Of course, if all the teams are using analytics, then you're right back where you started, because now it's about having the best analytics, or, more likely, getting lucky with your guesses.