EDIT: I think this line is the most telling:

> But he cautioned that the reverse mathematics approach may be most useful for revealing new connections between theorems that researchers have already proved. "It doesn’t tell us much, as far as we can say, about the complexity of statements which we do not know how to prove."

So, at this point, it helps us understand more about problems we already understand a little about, but nothing yet about new problems.