Accordingly, a realist account of the EU would stress that these states agreed to constrain their own autonomy and sovereignty largely in response to an unusual power configuration (i.e., the Cold War), and as much for security reasons as for purely economic ones. The end of the Cold War removed that power configuration, and we have seen the EU both expand and fray ever since. Germany's unwillingness to keep subsidizing profligate countries and European concerns about the implications of Germany's increasingly dominant role (as highlighted in this NYT article) are consistent with that view.
I think this is an especially poor account of both the history of the EU, and what an appropriate realist view of it could be. On the first point, the majority of significant EU integration has happened since the end of the Cold War. Prior to that it was almost exclusively and economic community, not a security community. (That's what NATO was for.) Since then the political and economic ties have grown stronger. True, the move towards the constitutionalization of the federation failed in 2004, but both Maastricht and Lisbon have happened since 1989, as has the Stability and Growth Pact and indeed the creation of the Euro itself. 15 of the 27 EU members have joined since the collapse of the Warsaw Pact. So there's been quite a lot of expansion in membership* and purpose. And to the extent that any fraying has gone on it's not because of power transitions or the dissolution of the USSR, but because the monetary union was built on the false premise that an accompanying fiscal union was unnecessary so long as countries pretended to abide by Maastricht. The fallout from the financial crisis has brought that illogic into sharp focus, but security concerns have absolutely nothing to do with it.
On the second point, it's right to emphasize Germany's dominant role, but wrong to think of it as only becoming salient recently. In fact, the best realist argument is the opposite: that the EU became more institutionalized post-Cold War -- including the establishment of the EMU -- in order to resolve the German question. The process of German reunification which, not coincidentally, corresponds to the period of increasing EU integration tilted power balance even further in Germany's direction. This also at a time when the US might have been assumed to draw down in Europe (Walt, like many others, expected and still expects NATO to dissolve any day now, or at least become a shell of its former self.) In other words, the EU serves a similar role with w/r/t Germany that NATO served with the US: by codifying and institutionalizing the relationships between more and less powerful states, the hope was that security dilemmas (including political economy security dilemmas) would dissipate. Or, as Ikenberry would put it, the EU integration project made Germany's commitment to "strategic restraint" credible.
I assume Walt would reject that argument, because a realist would (generally) argue that institutions are incapable of restraining great powers**, but if so then he has quite a lot left to explain.
*A lot of this is pure selection effect: many of the new entrants to the EU were part of the Warsaw Pact prior to 1991 and thus could not join the EU before. But that doesn't explain Austria, Finland, Sweden, and Cyprus. Nor Turkey's constant flirtations.
**This is why John Ikenberry isn't considered a realist, which is case-in-point of why realism is a dirty word these days.