You are correct. Political and bureaucratic types with no experience in building big systems called the shots until the 3rd week or so of October, and were making major change orders in August, and other changes in the week before launch. And no single manager had it as his personal, full time responsibility (although in fairness, it just occurred to me that anyone savvy in HHS would have avoided such a position like the plague). The infrastructure was grossly under-provisioned, and when the general contractor/integrator including integration testing, HHS's CMS (now "fired" and replaced by the private QSSI) finally did integration testing in the couple of weeks before launch it failed hard. But somehow, in best waterfall in practice tradition, that had no effect on the decision to launch.

As I like to say, no Google/Facebook/Lisper dream team could have possibly succeeded given the top down constraints. I can't even see a lot of the contractors putting in their best effort when they had to throw away so much work so often....