Nice articles here on the problems that led to the cancellation of the VentureStar, a single-stage to orbit vehicle:

Lockheed Martin X-33.
http://en.wikipedia.org/wiki/Lockheed_Martin_X-33

X-33/VentureStar - What really happened.
http://www.nasaspaceflight.com/content/?id=4180

Interestingly the main problem was making the liquid hydrogen tanks light enough, certainly not a high tech problem. I wonder if lightweight storage could be achieved by storing the hydrogen in very many micron-scale wide hollow fibers. See the table of tensile strengths here:

Tensile strength.
http://en.wikipedia.org/wiki/Tensile_strength

The solutions investigated for the hydrogen tanks for VentureStar included using high strength aluminum alloys or composite fiber tanks. The composite tanks were lighter but had a problem of debonding under high pressure. Note in the table of tensile strengths carbon fiber has a better strength to weight ratio than the aluminum alloy listed by a factor of 19 to 1. And the high strength glass fibers known as S-glass is better than the aluminum alloy by 10 to 1. There is also a special steel fiber known as scifer steel that is better than aluminum alloy by a factor of 4 to 1. It might even be for the carbon fibers and the S-glass fibers their strength to weight ratios are so high you wouldn't need to store the hydrogen in liquid form. You could store it as high density gas. That would eliminate the weight of the cryogenic systems for the hydrogen.
However, a key question here is whether this strength will be maintained in the radial direction. All the strengths listed for the fibers are for pulling along their lengths, i.e, their longitudinal tensile strength. But to use the fibers as thin hollow pressure tanks will require their strength to hold in the radial direction. After investigating this question before for hydrogen storage, I know that S-glass and scifer steel fibers do retain that strength in the radial directions. I'm not sure if this is true for the carbon fiber. BTW, the high strength polymer fibers in the table listed such as Kevlar, Dyneema, or Spectra are unsuitable because their strength only holds in the longitudinal direction, not radially.
Another key problem for using these fibers as hollow tubes is that they are only about 10 microns wide. So millions to billions of them would be needed to form sizable storage tanks. You would need a method of opening and closing these microscopically thin tubes at the same time for a throttleable engine. Perhaps one solution would be to have only a small portion of them being used at any one time and letting those completely empty out, then open another portion, and so on until all the fuel is used up. This would be an easier solution than having so many precisely controlled valves at the micro-scale that operated all in unison.