I think that it's tempting to write-off intel as a serious avocation because everyone is so wrong so often. But think about what you're up against? Not only is most of what you need to know in order to form opinions and come to conclusions actively hidden from you, but a big part of the hiding is also meant specifically to actively lead you astray. Predicting the future: it's hard. If intel was easy, there would be no need to do it. Of course in the case of Stratfor (though the government intel groups aren't immune to this), you have a product to sell and your ability to sell it is at least partially determined by whether you can convince your customers that you're confident of your work.
This is really no different from "ordinary" journalism; the same caveats and criticisms apply. There are lots of charlatans and blowhards out there. What's new?
/jordan