The missile defense test described in this Sep. 1 BBC report is certainly encouraging. The interceptor not only left the silo, but hit its target:
The Missile Defence Agency (MDA) said a dummy long-range missile launched from Alaska was hit by an interceptor sent up from California. ...[I]t has a mixed record, with only five successful tests out of nine.
During the last attempts to launch interceptor missiles in December 2004 and February 2005, the interceptors failed to launch from their silos.
I'm fairly confident that we could engineer and build a limited ballistic missile shield. Maybe a system robust enough to defend the country against 5 incoming missiles. Of course, no nation would launch 5 nuclear missiles at the United States. It would be suicide. The only response would be to retaliate immediately -- surely the President would not wait to see if the missiles could be intercepted. The launching nation would be vaporized. I suppose you could construct some scenario where the United States would choose not to respond -- say, terrorists seize a missile silo in a friendly country and somehow manage to launch a missile at the U.S. Such fantasy might make a decent Bond movie, but should not be the basis of defense policy.
The cost of the system is roughly $100 billion -- to date. If a workable system ever was developed, costs would skyrocket with actual deployment. One hundred billion dollars (if you keep Halliburton out of the mix) could build a lot of schools, hospitals, roads, and wells in the poorest parts of the world. Those parts of the world where hopelessness turns to rage and the disaffected become terrorists.