I've looked at the docs you linked to, and I don't see your point. ULA's document said that while the D-IV Heavy was in need of more work to make it single fault tolerant, they didn't state it was expensive or difficult to make those changes. There's a fault detection computer needing to be developed, and ground support stuff(crew access for example) too, but that wasn't presented as a challenge either. In general, the ULA doc adds fuel to the idea that man rating isn't a hard task in this case. The NASA funded study docs basically tossed out the idea of using D-IV Heavy as is, simply saying 'not human rated', and then went on to explore engine options and different second stages, mentioning again and again ways to increase commonality with Constellation hardware. The outright refusal to make use of the experience and common hardware of the current D-IV engines or existing second stage is a pretty baffling idea on the face of it, but once I remembered that protecting the workforce and contracts was a big part of Constellation's architecture, it made sense why they did that. That does mean this study started with biased assumptions, which means it's conclusions aren't useful to us; we're not hijacking D-IV Heavy to accommodate Constellation stuff, and it doesn't have much to say about using just a regular spec or minor upgrade of a D-IV heavy.
In my own opinion, I don't think there a big effort needed to man rate a EELV, at least nowhere near as needed as the ones for the Atlas I, Titan II, and Saturns of yore. There hasn't been as much hand wringing on an single issue today like Pogo was getting back then. At least, as far as I know.