The European unification of the post-WW2 era started out small, with some rather rerassuring cooperation. It quickly escalated and grabbed some low-hanging fruit: Tarriffs and visa requirements were removed. More cooperation followed, but the low-hanging fruits with great benefits at small costs were already gone.
European unification had turned into an ideology, though: To further European unification became considered a "pro" argument for new cooperation treaties or EC/EU actions in itself. it had become anend in itself, instead of a means.
Sooner or later some new European cooperation was bound to deliver less benefits than it incurred costs. I think the common currency is such a case* and the infamous bureaucratic excesses are other such cases.
I mentioned before how Niskanen's budget-maximising model of a bureaucracy explains such behaviour. Well, a different way to describe it is to apply the Peter Principle: Multinational cooperation (or bureaucracies) expand until they passed the optimum size. They may even expand to a point at which the net benefit is close to zero - especially if a step back is less possible than the clearly unsatisfactory total unravelling. The European Unification moved forward until it moved into territory which it couldn't master at all.
NATO has in my opinion shown similar behaviour, and I don't mean its expansion into East Europe:
The benefits of the classic defensive alliance were huge at least during the time of the Cold War. The Soviet Union with its satellites' auxiliary forces could have overwhelmed or isolated Western countries piecemeal if we hadn't stood together. The benefits of the new NATO playgrounds (military interventions, occupations and blue helmet missions) on the other hand are much less evident. In fact, the cost/benefit ratio may actually be horrible.
Still, reversing course is tricky (not as tricky as with the EU and Euro currency, though) and defendants quickly conjure the image of all or nothing; either play along fully or no alliance at all.
It seems to me as if a sternly modest approach towards bureaucracies and also multinational cooperation would be advisable. We shouldn't consider institutions as bad or evil; their net utility depends on their extent. A small institution (such as the original G6 meeting) may be spectacularly efficient and useful, while the same same in a less modest shape may be a waste of time (such as the Byzantinian G20 bureaucrat orgies) or even outright ruinous (as the Euro currency).