Interesting question. In my opinion a technology-based solution will never be effective. Every technology or system of layered technologies must always have gaps, for human safety of nothing else. These gaps will always be relatively easy to figure out from open-source data. This applies not just to aviation security but to a wide range of information that is kept "secret."

Law enforce is particularly bad at addressing terrorism. They like to pretend like they are preparing but almost every preparation is about response after the event.

Human intelligence could be effective in the situation you mentioned - domestic. I'd certainly work the intelligence side to try to narrow down information on the expected operation.

But fundamentally I think the most effective approach would be to develop an army of observers - business travelers especially - who report any strange behavior. Let's face it, it's not too hard to sit on a plane and "profile" 90% of the passengers around you.

I'm not saying it would be easy for an individual to accurately pick out the bad guys every time but if 3 or 4 observers identify the same suspicious person on the flight or in the gate area, it would seem like that could be enough to justify questioning by actual security people.