The international political system today is so much more complex in the scientific sense (interconnected, interdependent, co-evolving) that we have arguably (and it is the duty of political scientists not only to make the argument but to lay out the evidence, pro and con) reached a fundamentally new stage in which the cold-blooded "realist" rules of the past are, well, no longer realistic.
It may all come down to the ability to adapt, as in, "Remember the dinosaurs? Well, they couldn't." Everything functions as part of some system. You can choose which system to study (family, town, corporation, country, world; political, economic, cultural, environmental, biological), but you cannot omit the system and expect to understand human behavior. (In contrast, you can omit the system--i.e., the car--and understand the behavior of, say, a piston.)
Given that comprehension requires consideration of the system, one can start by trying to identify attributes of a system (e.g., defensive capacity) and examining each, perhaps by comparing cross-system functionality. That is feasible and has been done, but does not answer the mail because it is essentially a reductionist approach that still leaves you blind to how the system actually functions as a unit (you may imagine, if you insist, an intelligent supercar of the future in which the pistons were alive and constantly evolving to function better in response to whatever new fuel they were fed). The world's greatest expert on defensive capacity of nations would never understand nations without also knowing about other attributes, and I will be so bold as to claim (claims are not made to get hung up on; to "claim" is to invite refutation) that the most important of those attributes is adaptability. That is of course not a statement intended to be absolute - adaptability is key only in a rapidly changing environment. For neolithic man in a 90,000-year-long interglacial period, adaptability would perhaps not be particularly important. But for 21st century man--faced with population explosion, global warming, and asymmetric warfare--it is not much of an exaggeration to state that:
survival = F(adaptability).
The more complex the world becomes, the more opportunities will exist for finessing a situation. For example, I would hypothesize that:
Moreover, I would hypothesize that:H1 = the more complex the political context, the more opportunities a local rebellion will have to find allies to help it stave off defeat until it can figure out a winning strategy.
H2 = The more complex the political context, the more winning strategies there will be.
The new, more complex world of today facilitates the formation of new alliances that enable dissident groups to survive long enough to learn how to adapt. Once the process of adaptation gets rolling, it becomes increasingly difficult for the dominant group to keep up with the evolution of the adapting opponent: the dissident group not only can gain strength but evolve rapidly, creating new organizational structure, new behavior, and perhaps new goals, none of which are likely to be predicted by a dominant group. It will be hard for the dominant group to keep up because it will be reacting, always a step behind by definition. It will be hard because of a tricky characteristic of behavior in nonlinear situations (and much of behavior in a complex system will be nonlinear) - new behavior starts with almost invisible first steps and then rapidly snowballs. Disease control experts who worry about single cases of bird flu understand this dynamic. It will also be hard because we humans, and especially humans who are winning, tend to ignore the system, instead taking personal credit for victories that were probably mostly due to the luck of the draw.
It would seem likely increasingly to become a truism that today survival favors the adaptable. This may not have been the case for neolithic man or even for a Roman, but it does appear to be the case for, say, frogs (faced with environmental pressures that can be realistically be described as genocidal) and superpowers (Great Britain lasted a couple centuries as a superpower, the Soviet Union lasted 70 years, Nazi Germany lasted a decade or so, the U.S. looks shaky after some 60 years in that role and a short 17 years holding the position by itself).
Based on the above hypotheses, complexity provides more opportunities to learn a new, winning strategy. This does no good if you assume that you control the system, assume that you are a winner because you are superior, and assume that you will continue to be superior simply because you are at the moment. If, on the other hand, you are losing and therefore sufficiently concerned to open your eyes to the distinction between the influence of your own will and the influence of the system, then you can take advantage of these opportunities to learn faster than your smug opponent. Complexity provides both information from remote parts of the system to facilitate learning and opportunity to translate the learning into new behavior.
It is not just that hubris is dangerous - hubris is particularly dangerous in a rapidly evolving environment.