The X-33 was also supposed to dramatically reduce the cost of access to space. People had checked "the sums" Of course the problems started to arise when they actually started building the thing. Suddenly it wasn't as easy to build composite conformal LH2 tanks as they thought and the program went from "the SSTO future of cheap space travel", to a rusting curiosity. For Skylon to work, it requires everything to go just right, if 1 or 2 technologies don't pan out, or turn out to weigh more than originally expected, you suddenly have a vehicle that can't carry a useful payload. Every single SSTO project has turned out this way, brilliant on paper, but much harder once metal is bent. Maybe Skylon will be different, but they have a long, long, long road before the first one flies. I can understand the national pride angle (hey I'm Canadian, so I don't even have a bird in the fight), but I'm confident that Skylon will never fly.
SpaceX by luck or good planning has stumbled onto a virtuous cycle that puts them on the most viable path to reliability and to reducing the cost of access to space. How:
1. They built a heavily over-engineered largely conventional rocket that has a viable path towards reusability. They can build a little, test a little, get missions under their belt and learn how the Falcon 9 flies and where they can optimize performance (Falcon 9 "full thrust" launches later this year). However even if all these reusability experiments fail for some reason, they still have a profitable launch vehicle. If Skylon fails, they've got nothing, except maybe a very very expensive ballistic transport.
2. All of the R&D flights are on someone elses dime. By building the Falcon 9 larger than what is needed for most LEO payloads, it can fly those payloads at a profit, and then, once the second stage is away, do what ever experiments they wish to test out various recovery options. They are running a viable business while building up the experience and expertise they need to reuse a stage. Most launch vehicles are built to the smallest possible size and then scaled with SRBs so any flight test incurs additional costs (even if its just for the SRBs). Skylon requires $12 billion up front to build a vehicle, someone needs to be willing to invest that money on a vehicle that may or may not work.
3. Development costs, to 2014 SpaceX has invested ~$1 billion to develop both Dragon and Falcon 9 (both their money and NASA's). The initial Falcon 9 version took $300 million to develop (less than what was spent to just develop Skylon's heat exchanger). For argument sake, lets say it will cost SpaceX $1 billion in R&D on the Falcon 9 before they can start recovering stages, and over the lifetime of the vehicle they fly 200 times. That's $5 million per flight in R&D costs. Skylon at $12 billion (and those were 2004 numbers), even it if flies 1000 times, is $12 million a flight in amortized R&D costs per flight (almost a quarter of the price of an expendable Falcon 9). Skylon estimates amortized launch costs at $1000/kg, SpaceX is hoping for $1100/kg, so while operational costs of Skylon may theoretically be much cheaper. when you account for the R&D costs, the differences are negligible (and Falcon 9 is far lower risk than Skylon)
4. Scaleability. To scale Falcon 9, you can stretch the tanks, upgrade the engines, or in the case of Falcon Heavy, stack cores . With the exception of Falcon Heavy, all of these have already been demonstrated. To scale Skylon requires the entire vehicle to be redesigned.
In 3 years when REL starts looking for investors to fund the rest of the development of Skylon (even if they now have a demonstrable engine), are those dollars going to flow to a high risk, high cost proposal, or to SpaceX who by routinely recovering and reusing stages has just quartered the cost of launch