I'm not quite sure I'd put it that way. I do think that the risk of most cargo is overstated here. A very large amount of air cargo are shipments between known shippers. If a Boeing plant in Seattle orders a few crates of ball-bearings from a subcontractor in Buffalo, the chances that terrorists would be able to sneak a bomb into that shipment is quite low. Not zero, but low. Significantly lower than the chances of them being able to do so on the person of a passenger.

In my mind, the fundamental difference between cargo and passengers is that it's not considered acceptable (for a number of reasons) to divide passengers into categories based on risk. But it is for cargo (e.g., the example above). So a methodology that concentrates mostly on the higher-risk cargo seems appropriate to me: it doesn't bother me that much that we don't screen 100% of all cargo.