Costs are from the space X Mars presentation.
http://www.spacex.com/mars
I did make a small mistake as it'd be 4.26m for passengers and 3.4m for cargo, so around $17000 for a 250 person launch.
Costs on page 41 for Mars transit, which is 6 launches. Spaceship costs per Mars launch are higher than the tanker launch because it's stuck on Mars for two years, limiting it's reuse. It would have 100 launches like the tanker if used to LEO.
Booster per launch
11 / 6 = 1.8m
Spaceship per launch (scaled tanker cost, as other than seats and the bar they're the same vehicle)
(8 / 6) * (200/130) = 2.46m
Passenger cost
1.8 + 2.46 = 4.26m
4.26m / 250 folks = $17,046
Think that's all in order! Phone screen and calculator isn't ideal.
Next presentation is in a couple of months. Tweet from Elon suggest they've now sorted out their funding beyond stealing underpants.
I did mean to say that it would still stand that something like StarTram is probably less noisy and could achieve a high launch cadance than big reusables. You'd need 16000 rocket launches over a decade to put 4m people in orbit.
That's why I think rockets can establish the market. Investment in something like StarTram can then make more sense as a 'railroad' investment serving that market. Assuming gen2 big boosters don't cost it out.
As another aside, if SpaceX fail for some reason, Bezos is investing $1bn a year in Blue Origin by selling Amazon shares. Big reusables will happen on way or another.