Roads are a high-fixed-cost-low-variable-cost industry. I posted on these at Reddit - http://reddit.com/info/2fquz/comments/c2fttk. Everything that's wrong with airlines will go wrong with roads, and more. I suspect that you'll see massive overinvestment in roads as road companies try to capture market share, then the huge number of intersections (all toll, remember?) will reduce driving efficiency.
Free markets solve a lot of things, but they don't solve everything. There's no magic wand that makes them more efficient than a public solution. There are, however, a series of incentives and information-transmission mechanisms that usually give a free-market operator an information advantage over a public operator, hence letting them produce more efficiently. If the incentives point in the wrong direction, though, you get less efficient production.
Similarly for another high fixed cost industry which you cite - telecommunications.
Free markets might not solve everything, but they're almost always better than the alternatives. As examples of the alternatives, consider the postal service, government-run schools, Medicare and welfare.
You may currently be spending more than $1400 to the government to pay for government roads. With competitive incentives in road placement and size, you might save more than enough time to offset the transaction costs.