Brownian Emotion has one of the best discussions I have seen on the tension between national security and an open society with respect to Google Earth in particular, and spatial data in general. http://www.brownianemotion.org/2007/06/29/google-earth-anxiety/. It is dialogue between Assemblyman Michael Gianaris' concerns about Google Earth and Avi Bar-Ze'ev, one of the forces behind Google Earth technology. Although I think that at times the two are speaking past one another, they both make some excellent points.
I am somewhat conflicted on this topic, as I can see both sides. As a former intelligence analyst -- including a satellite imagery analyst - I believe that spatial data, such as commercial satellite imagery, can have some real value to terrorists. And I do not fully buy into the argument that they can simply get the data elsewhere. Even if true, each additional step that they are forced to take increases the risk that they will either get caught before they are able to cause harm, or that they leave a trail that can be traced back to them. Law enforcement officials recognized this with respect to the flow of money to support terrorism and now banking transactions are monitored closely.
And as spatial technology improves (such as 3-D visualization), potential terrorists and their allies are going to be able to do more of their planning over the Internet. This will make it even more difficult to monitor them. Some have argued that such training will not provide a sufficient level of comfort to prepare for a terrorist attack - the spatial data may not be current, or may not provide all the necessary information. However, it appears that terrorists are not looking for carefully planned, precision strikes, but simply to maximize casualties and terror. (For example, it does not seem that are looking for a means of escape). As as a result, mid-quality, older data may be sufficient.
On the other hand, I agree that this industry has come too far and has too many potential economic and societal benefits to be hamstrung by censorship or stringent regulation. Unfortunately, there is no legal regime in place to balance these competing issues with respect to spatial data. I hope that a useful dialogue will develop soon between the spatial technology industry and the US intelligence and defense communities. Otherwise poorly thought through legislation may hamper the development of this segment of the industry. Perhaps the correspondence cited above will be a step towards such dialogue.