The curmudgeon of the space pundit community, Jeffrey Bell, has an intriguing screed against Single Stage to Orbit
rockets that deserves a look, followed by an intellectual discussion.
It should be noted that only a very small minority of the engineering community believes in single-stage rockets. All competent engineers know that, while the basic governing equations of space flight do not exclude Single-Stage To Orbit (SSTO) reusable rockets, the equations do tell us that SSTO is enormously challenging and downright impractical.
The Delta Clipper (DC-X) has often been dismissed as a stunt by supporters of the conventional "expendible rockets" paradigm. While DC-X clearly did not get us any closer to a single-stage reusable rocket (contrary to the optimism of G. Harry Stein in the recommended Single Stage to Anywhere
,) the DC-X was an important program that did make several valuable contributions. It showed that a hydrogen-fueled rocket could be launched from austere facilities by a skeleton crew, hover and land, then get turned around and reflown in a reasonable amount of time. The project was an example of lean management and rapid prototyping.
If Delta Clipper had been fully funded, perhaps it could have resulted in a reusable SSTO. The concept was to build and test (now given the buzzword nickname of "spiral development,") incrimentally achieving the full performance after a series of increasingly-capable prototypes. In the mid-90's Mitchell Burnside Clapp (at the time, a Captain at Kirtland) proposed building a Mach-3 capable "DC-XB" and an enlarged "DC-XC."
In short, while Jeff Bell's physics and his mass figures for DC-X are indisputable, it's not clear that DC-X was the pinnacle of SSTO technology. In actuality, DC-X was probably on the heavy side, compared to what McDonnell Douglas could have built with cutting-edge materials and structures. DC-X was designed for a simple mission: take off, hover, and land. A high fuel fraction isn't needed for that.
Assuming that we could have incrimentally built ourselves up to SSTO (perhaps by the time "DC-XZ" came along,) the useful payload would have been such a small fraction of the total launch mass that we'd have to ask if all the effort was worth it. An SSTO will always have to carry the mass of empty tankage to orbit. A Two-Stage To Orbit (TSTO) allows you to drop off some of the dead weight, recover it for reuse, and put more payload in orbit. The more stages you have, the more payload in orbit for a given liftoff weight. However, the addition of each stage creates more complexity, and each staging event carries a significant risk of ending the mission.
After taking a long look at the old concepts for reusable launch vehicles, I have to favor something akin to Max Faget's "DC-3" space shuttle concept. The two-stage rocket (fully reusable, unlike the space shuttle that emerged) consisted of a booster and orbiter that both resembled airplanes. The booster would burn out, then return to base with its onboard jet engines. The upper stage (orbiter) would only ignite after the booster burned out, and it would fly to orbit without dropping propellant tanks or solid rocket boosters.
The critical flaw in Faget's DC-3 was that the orbiter dropped like a brick during re-entry, from the entry interface all the way until it reached subsonic speeds. This was a great technique for reducing vehicle heating (which is why SpaceShipOne uses a variation of this re-entry mode,) but it introduces a tendency to spin at hypersonic speeds.
Faget's orbiter had some novel features, like a straight, high aspect-ratio wing for good subsonic flight characteristics, and jet engines to enable ferry flights and aborted landings. However, it would be a shrinking NASA budget (and a demand from the Nixon White House that NASA's shuttle conform to the Air Force's proposed Vandenberg mission requirements) that doomed fully-reusable shuttles like Max Faget's DC-3. We got stuck with the space jalopy we have now, and the safety of the astronauts was the tradeoff.
If Burt Rutan or other rocket visionaries want to build a reusable orbital rocket, they would be wise to look at the work performed during the late 60's, before the shuttle's development budget was cut. They would also be wise to build a mini-shuttle (akin to the X-20) that would take many of the unknowns out of development.
The space shuttle's biggest failing is that it tested too many new technologies despite being an "operational" system (opposed to being a prototype system.) No vehicle had ever flown with the shuttle's fragile heat shield or reusable hydrogen engines before. As a result, NASA had no clue how difficult it would be to maintain the orbiters between flights until it had actually flown them a few times.
With hindsight being 20/20, we should have a good idea of how to competently build a reusable rocket. The question is, who will build it, what will its mission be, and where will they find the startup funds?