>The problem with this idea is that it rests on two incorrect assumptions - that there is some governing body that would be able to prevent the release of such information, and that we would immediately know upon detection that the asteroid was on a collision course.

Not necessarily. Only a couple people knew about this one before it had whizzed by. While I agree it'd be impossible to keep an upcoming impact secret, if the timeframe is short enough it would only take a few select people to keep it contained long enough to no longer matter.

And choosing to keep it secret wouldn't only be done if collision was known to be certain. Just having a moderate chance of impact might be enough.