Roentgens? That's like measuring fuel flow in hogsheads per fortnight. (All due deference to you folks in the U.K., but "fortnight" to us in the U.S. seems appropriately archaic.) And yeah, that probably means he's following the Blunder's lead and relying on stuff from the late 1950s.
In all fairness even fair-minded, intelligent people have trouble with the notions of flux and energy, of the Van Allen belts as exhibit widely different measurements of these values based on your location and the solar weather, and of integrated dose as being highly dependent on one's path through three-dimensional space. Several times I've been asked for "general" figures for "How much radiation is there in the Van Allen belts?" and I do my best to decline to answer, knowing that whatever qualifications I try to apply to an answer will be omitted. "Jay says there are blah-blah rads..." probably still pops up from time to time.
But that's no excuse for Adrian's consistent and hopeless inability to understand that he is conceptually wrong on nearly every count, on nearly every question he looks at. He can't be made to understand that he doesn't understand. He desperately wants a concession on "the data are correct" without being willing to consider that that the data are inapplicable to the problem at hand. To him, "Can we agree that the data are correct?" is indistinguishable from questions of applicability. To him, "correct" includes "applicable," and it's a massive paradigm shift for him to believe that those may be two separate concepts.
It's really no different than his misunderstanding of the Saturn V issue. Once he gets it in his head that "Jay admits the plans were lost," there is no prying him loose from his simplistic understanding of how aerospace designs really are documented. In his mind any concession, no matter how carefully qualified or explained, validates his simplistic misunderstandings.