My guess about the official's initial "I don't know why" reaction is that the official didn't realize (until asked) that the software had counted things this way and was caught off guard by the question.
I wouldn't expect any further official explanation, because the data explains itself - the first pages of the data are clearly calculated in terms of "Cards Cast", each ballot clearly consisted of two cards, the remainder of the data clearly lists the number of ballots cast in each race (which clearly shows a ~70% turnout), and the official state-level reporting clearly lists a ~70% turnout. To put it bluntly, I wouldn't expect any further official explanation because the data clearly explains itself to anyone who takes the time to actually read past the first two pages.
C of C
I took the time.
Are you sure you don’t know why Muffle Oar outed you?
See my last post. For instance, how do you get an odd number of cards cast in your scenario? And why does subtracting the overvotes (disqualified Presidential votes) and undervotes (no Presidential vote) result in less Presidential votes (using your method) than the number of Presidential votes they say were cast?