You might be surprised to hear me talk about a solution to issues like this from a perspective of data nationalism, being a GNU software user and Free Software Foundation supporter. However, I am also about solutions to problems, and advanced forms of data nationalism are a direct approach to ending these types of network attacks and remote data breeches that cross continents.
The "world wide" web as we know it has reached its end, anyway. It's time, as security analysts, to put on the tinfoil hats and take a few doses of paranoia. First, access to the Internet is a service already, so start treating it more like one. Federate the Internet (US, China, Canada, Brazil, etc.) and write interfaces between each unique instance of Internet ecosystems that cost money to access; expensive access, at that.
Once global networks and Internet providers are fractured and new standards and protocols are put in place to keep everyone with once open access out, hacking threats from other countries will initially be zero. Of course, over time, the same hackers we worried about before will figure out ways to get in; however, now the pipe will be a single entry point and a small one, at that. We can more easily monitor and prevent intrusion from non-USA would-be hackers.
Companies like Google and Facebook - who federated Internet supporters initially worried would never support such initiatives - could actually stand to benefit greatly from such changes, being able to demand high costs of countries like China and the EU whose users will want access to American Facebook and G+ users.
It's another one of those tough decisions and unpopular ideas that holds incredible opportunity for control and security overall for American Internet-based companies and Government agencies and resources, but whose implementation just plain scares too many people. It may be time to get over the fear.