Yeah. But as Scott Aaronson points out, GI has a very different feel to factoring. A randomly chosen number will be hard to factor, but it's actually quite hard to come up with examples of GI that are hard:
> But then again, in practice, graph isomorphism has already been “basically in P” for decades! If you have two large graphs for which you actually need to know whether they’re isomorphic, just download NAUTY and run it.
> This contrasts with the case of factoring, for which I’d personally say that it remains much less clear whether it should or shouldn’t be in P.