Back of the envelope calculation for the price of an Earth-to-Earth ticket:
Musk's stated goal is $500k/ticket to Mars.
It's a shorter trip, so perhaps ~5x as many passengers in the same volume (i.e. 500 total; cf. a380 which seats 850).
It takes five (?) orbital refuel trips for the martian journey, but we'd need none of those. Depending on how much less than a full tank the passenger vehicle needs (payload could be smaller; the ship would also not reach fully orbital velocity), the fuel cost would be between 1/5 and 1/10 the Mars fuel cost.
So that would bring the cost down to between $10 and $20k/ticket, within reach of business travelers.
If maximum the number of flight cycles per vehicle is greater for Earth-to-Earth trips than for Mars, then that could further reduce the ticket cost. It's unclear to me which direction that number would go-- Earth's atmosphere is much thicker on re-entry, though the velocities will be much lower than an interplanetary re-entry. Since aerodynamic drag goes as the cube of velocity but only linearly with density, I'm guessing the speed would matter far more. That would imply much better lifetime on Earth.
So if the E2E fuselage gets (conservatively) only 2x as many flight cycles as a Mars trip, that could bring down the per-seat cost to $5k-- now getting close to the cost of an ordinary international ticket. Of course this is all assuming that Musk's baseline of $500k to Mars is reasonable.
Would be curious to hear from some rocket engineers about these guesses at the numbers/efficiency.
Not accounted for is amortized development cost for E2E-only vehicles, as well as all the infrastructure and ground support at the destinations.
Edit: If you wanted to be less conservative, you could pack in 1000 people instead of 500 (0.5x ticket price multiplier), or use a different source for the Mars ticket price (0.4x), which would bring it to $1k.