I don't know anything about quantum computing, that said however, you are misunderstanding me because I didn't explain it adequately.
Even if you are talking about something like spin, which is either up or down, it still doesn't work. There is infinite uncertainty in the state of the particle before you make a measurement, and you cannot conclude that what you observed is different form it's previous state. Although your observation does cause the superposition to collapse, there is no way to predict which state it will collapse into. if you could do that the uncertainty would be meaningless. The history of the particle has no effect on the future outcome. It's like flipping a coin. While it is spinning in the air there is a great deal of uncertainty which side is up. as soon as you catch it to find out, this "superposition" collapses into one of two states, but just because it was tails last time does not imply that it will be heads this time. Over a long series of measurements you'd be right: it lands tails just as often as it lands heads, but that has no bearing on the outcome of an individual toss.
I think you are right that quantum computing relies on entanglement in some way, but I think the difference is that somehow the observation step is removed. I would also venture to guess that there is some quirk of the process that limits everything to light-speed.