The decline of the nation-state is seen in a graph of the ability of small groups to replicate the state's most vital commodity -- large scale violence. 
The Yale economist, Martin Shubik examines this in his paper "Terrorism, Technology, and the Socioeconomics of Death" (PDF). His conclusion? Rapid technological improvement and global information transfer (part of a larger context of interconnectivity) has produced a spike in the ability of small groups to produce mass casualties (see attached graph).
Loose Nukes, Biological weapons, and Traditional Terrorism
Armed with weapons of mass destruction, this conclusion is certainly true. Small groups armed with these weapons would be able to approximate the level of violence of nation-
states. Fortunately, there are caveats. Shubik's economic analysis indicates that nuclear weapons are particularly difficult to produce, acquire, store, and deploy -- so much so that it may exceed the capabilities of small terrorist groups. The lower technical hurdles of biological and chemical weapons represent more of a threat. Regardless, the current complexity of these weapons implies that their use will be relatively infrequent. Over the longer term this will likely change. The rapid technological improvements underway in the biological sciences will eventually lower barriers to entry and thereby increase the potential of use.
Frequency of Large Scale Attacks
In the short to medium term, even with the availability of chemical and biological weapons, we can expect large mass casualty events to be relatively rare. How rare? Clauset and Young answer this question in their paper "Scale Invariance in Global Terrorism" (PDF). They found, from the analysis of data from 36 years of terrorist attacks, that casualties per terrorist attack follow a power law (scale free).
The distribution of attacks shows that most terrorist attacks generate few casualties and only rarely do attacks produce large numbers of casualties. This analysis demonstrates the difficultly terrorism has generating large casualty events. In fact, their analysis indicates that an event larger than 9/11 is only likely in the next 7 years. This may seem horrible, but as a challenge to nation-state military power, this falls well short of a transformative capability.
System Disruption and the Democratization of Violence
If we look at different metrics of violence, such as the economic costs of system disruption, the picture changes dramatically. Unlike traditional terrorism, system disruption doesn't focus on casualties but rather on the dislocation of infrastructures and markets. The effectiveness of these attacks are measured in the financial damage it causes the target economies.
Analysis indicates that the results of attacks that cause system disruption do not follow a power law but rather a linear function. This makes the method much more suitable for sustained warfare against nation-state targets. Attacks can be planned with a relatively high degree of confidence in the results. Additionally, the results are sufficient to provide substantial returns on the invested effort and capital (direct losses to Iraq due to systems attacks are over $7 billion, to the world economy the damage is in the hundreds of billions due to the influence of the attacks on the supply of oil to global markets). The reasons for this superior performance include:
- The barriers to systems disruption are de minimus. The methods are therefore available to the vast majority of groups that attempt it (a 99.9% solution). Specialized knowledge helps, but it isn't necessary to accomplish an attack with a substantial impact.
- Infrastructures and markets provide vast vulnerabilities that can be exploited with relative safety. As a result, attacks against systems can be easily replicated over time -- for example, routine attacks on gas and oil pipelines that connect to the Iraqi refinery/power plant complex in Baiji usually result in $50 million + in damage per attack.
- Attacks gain leverage from the technology and interconnectedness of the networks being attacked. Even small attacks can generate outsized returns. In contrast to traditional terrorism, systems attacks do not suffer diminishing returns.
The quantity of damage routinely generated by systems disruption far exceeds the pay-off of traditional terrorism (the area under the curves). This technique is therefore a viable method of warfare that can challenge nation-state military power today.