While I don't question their value, I don't know that analytics will ever have the predictive abilities in football that they do in baseball. The sample sizes are always much, much smaller.
I can agree with that. At the same time, I think that the concepts learned from baseball can be applied to other sports in a completely different way. They definitely serve a different role in the NFL than they did/do in MLB but are equally useful in the right contexts. For example, financial metrics are probably more important than game stats when you're underneath a salary cap.
I think there are probably a lot of really interesting, complex ways to pick apart the draft and how it relates to a team's monetary resources/draft picks that could help teams become more efficient at drafting and managing the salary cap, at least in terms of rookie production versus expensive free agent contracts. Sure, everyone knows this on an intuitive level, but what if you can get it down to a science? Think about situational football. Good coaches instinctively known when to go for two in the fourth quarter, but someone with a lot of brain power put it all together in a pretty fool-proof chart -- that's basic addition compared to what's possible with the right mental resources at work. Remember when the Patriots decided that it was more economical to stock up on second round picks than any other round? It didn't work, but at least they put together a testable model. Eventually, someone is going to get it right, and everyone else is going to be playing catch-up trying to imitate it.