The NORAD testimony I was noting is in fact contradicted by these NORAD tapes. To quote Vanity Fair's piece again:

"In the chronology presented to the 9/11 commission, Colonel Scott put the time NORAD was first notified about United 93 at 9:16 a.m., from which time, he said, commanders tracked the flight closely. (It crashed at 10:03 a.m.) If it had indeed been necessary to "take lives in the air" with United 93, or any incoming flight to Washington, the two armed fighters from Langley Air Force Base in Virginia would have been the ones called upon to carry out the shootdown. In Colonel Scott's account, those jets were given the order to launch at 9:24, within seconds of NEADS's receiving the F.A.A.'s report of the possible hijacking of American 77, the plane that would ultimately hit the Pentagon. This time line suggests the system was starting to work: the F.A.A. reports a hijacking, and the military reacts instantaneously. Launching after the report of American 77 would, in theory, have put the fighters in the air and in position over Washington in plenty of time to react to United 93."

But it didn't happen that way, and the two fighters that were over Washington that would have taken down United 93 were unarmed after recent training missions. If they had taken down the plane, it would not have been a shootdown, but a kamikaze mission: Marc H. Sasseville would take out the cockpit and Heather Penney would take out the tail, on the premise that just taking out an engine could leave them with a functioning glider.

> The only gap I could fault all involved with would be somehow expecting the unexpected -- that plane hijackers wouldn't be interested in flying the plane to safety, but would instead intentionally fly it into a building.

Sadly it really shouldn't have been unexpected.

> ...between 1994 and 2001 CTC [CIA Counter Terrorism Center] collated no less than twelve specific reports warning that terrorists were scheming to hijack an airliner and fly it into a prestige target. Several of the reports actually named bin Laden and al Qai'da..

(from the book Military Intelligence Blunders and Cover-Ups by John Hughes-Wilson)

The amount of intelligence collected really shouldn't have been missed. There was tons of it, from all sorts of sources, including suspected terrorists on lists that were taking flying courses and asking about flying big airliners, who didn't have any interest in landing or taking-off It was an epic failure on the scale of the US intelligence epic failure of Pearl Harbor.

12 reports in 7 years?

This is the needle in a haystack of needles problem. Of course it's obvious in hindsight.

But cueing the entire civilian ATC / ANG system to potentially shoot airliners out of the sky within 30 minutes?

Expecting that to be in place, even with overwhelming in number but unrealized reports, is unrealistic.

> 12 reports in 7 years?

Reports on new capabilities and tactics of a terrorist organisation that has struck US targets and is considered among the most dangerous?

And then you have concrete reports on the ground of terrorist-connected people taking flying lessons and asking about bigger planes.

Even without hindsight that's pretty obvious things to relate and take note of.

That's making the mistake of assuming these reports were unique in some way, as opposed to being part of a stream of thousands of other reports, some including even more threatening implications.

In low signal, high noise scenarios, the hindsight bias is not saying "We should have known this meant something," but rather saying "We should have given this particular thing more attention."

In our world, it happens with logs. Post- or during an incident, reading back through a log to find the culprit is trivial.

But how easy is it to have a log continually dump to your terminal, all day, for weeks, months, years, and identify the root cause line as it happens and before other monitoring tripwires go off?