There is a good analysis of the reusable rockets vs. space elevator debate by Jordan Kare over at
http://web.archive.org/web/20031204105315/http://www.isr.us/spaceelevatorconference/presentations.html
http://web.archive.org/web/20031204105315/http://www.isr.us/spaceelevatorconference/pdf/Kare/Workshop2_kare.pdf
It is completely reasonable that a fully reusable LOX/hydrocarbon "space truck" - using off-the-shelf engines like the NK43 and RD-0124 - could deliver payloads to LEO for about $50,000 per ton, and could fly as often as eight times a day.
There is a good opinion of why the costs of reaching orbit have not gone down after almost 50 years over at:
http://groups.google.com/group/sci.space.policy/msg/72844bd6b7977af3
Here is the important quote from the usenet posting:
Because, by the nature of the early adopters of rocketry, the people historically in the business of rocket-building have found it most profitable to adopt the business model of Rolls Royce.
No, worse than that. Rolls Royce directly benefits if it can reduce the actual *cost* of a car, so long as it does not reduce performance or prestige or any of the other things that let it charge a high *price* for the machine. So it will quietly pursue modest cost reduction, and pocket the difference as profit, and maybe we can hope for a price war between Rolls and Bentley some time down the road.
The major customers for rocketry mandate a perverse pricing structure in which every reduction in cost translates directly to a *reduction* in profit for the builder. Rocket builders serving traditional markets benefit from building the most expensive rockets they can get away with.
Rocket builders trying to pursue cost reductions and low-cost business strategies, are trying to do something that has never been done before in spite of people arguing for its desirability, and find their efforts confounded not just by the actual difficulty of doing something new, but by foolish arguments of the form, "X hasn't been done before in spite of arguments for its desirability, therefore there must be some reason why X cannot be done..."