Missing the systemic forest for the unipolar trees

I've read (rather quickly) the exchange between J. Busby (at Duck of Minerva) and N. Monteiro about the latter's recent article in International Security, which argues that, to quote Monteiro's blog post, "the conventional view (largely based on Wohlforth’s work) that unipolarity is peaceful misses some important conflict-producing mechanisms particular to a unipolar world."

Later in the same post, Monteiro writes:

My argument that unipolarity makes room for particular conflict-producing mechanisms — involving the unipole in case it does not disengage from the world — is indeed compatible with different assessments of the overall level of conflict in the system, as well as of the lethality of those wars.

Let's assume for the sake of argument, as the Busby/Monteiro exchange assumes, that the system is currently unipolar (this is a definitional question and there is a case to be made that the system is not unipolar, but we'll set that aside). Here's the situation (at least as it appears to me): the period since the onset of this assumed unipolarity (i.e., since the end of bipolarity with the end of the Cold War) has been unusually peaceful for the system as a whole, but not for the 'unipole' (i.e., the U.S.) itself. Monteiro's empirical focus, from what I gather, is the second part of this situation -- the wars the U.S. has been involved in recently -- rather than the first part, namely the level of violence in the system as a whole. But it's the first part that's more important.

In other words, if the current system is trending in a peaceful direction, as seems to be the case (see J. Goldstein's recent book, which I will be posting a fairly long review of here soon), then it becomes somewhat irrelevant that unipolarity may "make room for...conflict-producing mechanisms." In a system that, for reasons other than polarity, is becoming more peaceful, those "conflict-producing mechanisms" are just not going to be producing much conflict.