It's impractical, though. Building a wheeled vehicle capable of reaching such speeds is more technically challenging than building a simple LEM-style rocket ship. Also, as the velocity increases and the vehicle approaches orbital speeds, the traction will decrease since it is a function of weight on the road surface, and you will likely see wheel slippage.
It doesn't need to be a wheeled vehicle. Maglev would work, and since it doesn't rely on ground contact, traction wouldn't be an issue.
Since the vehicle has to have rockets on it to complete the insertion and for orbital maneuvering, why bother with wheels and traction and motor drive trains? A much lighter pure rocket ship makes more sense.
The advantage to maglev (or even to wheels in the ideal case with no nasty obstacles like friction and frictional heating) is that whatever object you're taking off from serves as reaction mass and your delta-V at any point is exactly equal to your forward velocity, which means that you don't have to carry your reaction mass (except for your apoapsis maneuver) and that your energy efficiency is maximized. Also in the case of a "train" configuration (such as maglev) rather than a "car" (such as you might have with an idealized, frictionless, wheeled vehicle), the vehicle can be externally powered, allowing you to get your launch energy from the local power grid instead of having to carry your energy source with you.
On an airless body, I imagine maglev launch would be quite competitive with rocket launch, assuming you could ever get enough of a colony going to justify the infrastructure investment. Of course, the difference from palebluevoice's scheme is that, early-flight terrain clearance permitting, you'd want your launch to be completely in the horizontal.
If Earth were an airless body (and, magically, could still could support human civilization) it's a good bet that we'd have something like that today (both for really fast surface transport and for space launch). However, to get the maglev track above the atmosphere on the real Earth, we'd need something like a [ame="http://en.wikipedia.org/wiki/Lofstrom_loop"]Lofstrom Loop[/ame], which makes both the technological and the economic requirements much more daunting.
I think it would work better with a system similar to those upside down rollercoasters, where the load is hanging on a rail. Scale that up, and you could put a cubesat, a satellite, a Deltaglider, whatever there.
