I don't think $1.4 billion is the worst-case cost, I think that's the best case. Given that the two 5 segment SRB's are going to run about $160 million, the four RS-25 D or E's are going to cost $200 million (which is highly optimistic), the J-2X's or RL-10's will cost about $70 million, the tanks will cost about $85 million, and the Orion costs about $450 million, you're looking at close to a billion a launch even if the ground requirements were just three guys in a Walmart parking lot and liquid hydrogen was free. At two flights per year the annual cost would be over $4 billion, but the most optimistic budgeting is for about one flight every two years.
Some at NASA actually like my suggestion of putting NASA ground personnel and their families into cryogenic suspension so they don't suffer skill degradation in between launches. ^_^
Elon will probably give up on hyperloop pretty quickly. I roughly calculated the allowable tube deviations and they're quite small (an inch or two causes issues at transonic speeds). I've ridden on an Amtrack route where my head kept rhythmically slamming into the wall, hard, due to track discontinuities. The side G-loads from a given deviation goes up with the square of the vehicle's velocity, so what was bad at 70 mph would be a hundred times worse at 700 mph.