NASA's dream of a space plane Shuttle replacement capable of carrying a 40,000 pound payload into orbit will almost certainly end up (if it's ever built) being just as big, just as complex, just as expensive as the Shuttle itself.

Aerospace vehicles cost by the pound, the the pounds snowball quickly. For each additional pound you want to put into orbit, you need several additional pounds of propellant. You also need bigger tanks, heavier structure, bigger engines, which need more propellant, and so on. Ligher materials and more efficient engines can get you part of the way, but the engines on the Shuttle are already pretty damned efficient and the materials are already pretty light. You can gain some at the margins, but there is no order of magnitude improvement on the horizon.

A manned spaceplane capable of carrying a crew of 7 and perhaps 5,000 pounds of supplies to the ISS could be built with a takeoff weight of perhaps 60,000 pounds verses the 200,000+ pounds loaded weight for a Shuttle. Scaling down, you could then have a booster that's a fraction of the size and cost of the current Shuttle ET and SRBs.

As for large payloads, very few of them need humans in space. Expendable boosters on the order of the classic Big Dumb Boosters can be made a lot cheaper than human rated systems. The Russian Soyuz booster has flown some 2000 times (if you count the Molynia booster, which uses the Soyuz booster with an additional upper stage) with a high reliability rate. It can already meet the desired $1000 per pound to LEO that we'd need to spend billions to achieve. The Proton is even more economical. Sure, the Russians have many advantages in that regard, such as very low R&D costs and labor rates.

Probably the biggest reason why American and European boosters are still so expensive is that there is little financial incentive to develop cheaper ones. The Russians are limited in how cheap they can sell their boosters so they won't undermine our launch industry, so they don't mind getting paid $30 million for a flight that cost then $15 million (or less). The American booster companies have made some improvements along the margins with the ELV (heavily subsidized R&D by the DoD), but they have no incentive to go for really inexpensive boosters.

Perhaps what is needed is for the DoD to say that, after a certain date, say 2005, they won't pay more than X dollars per pound to LEO. Let the manufacturers really compete for the launch business. I've seen too many US companies gold plate their designs rather than concentrate on basic functionality, economy, and reliability.