I would prefer the cheaper and more effective way; profiling.
Also, you can't say security has been working well-- look at the number of incidences of things going through security accidentally via negligence (knives, guns, etc)-- while there's no official numbers, the anecdotal evidence is quite moving.
Actually, there
is documented evidence (which I'm not going to look up, because it supports your contention). The TSA does publish numbers (though buried deep in their reports) on the number of times undercover agents are able to slip weapons through security on training/testing runs. The number is quite high, if you look at it in a "Sky is falling way". But that is the incomplete picture.
Suppose, just for argument's sake, you actually have a 50/50 chance of slipping something through security. Is that "good enough" to mount an operation? Consider that there are at least a dozen people involved, to support just one operative. You can try to separate them into cells - but that doesn't mean that they are entirely hidden... it just gives them time to try to escape while their links are followed. Plus, there is a lot of money involved.
Do you risk those 12 people, plus a large chunk of scarce resources, on a venture that only has a 50/50 chance of getting something onto the plane. (we haven't even considered that most bombs on planes lately have not gone off properly, eg. shoe bomber and underwear bomber)... or that if the intent is to forcibly take over the plane there
might be sky marshall - or just a plane load of passengers who are not going to sit idly by.
So you try and reduce that risk by making the plan more "fool proof" and sophisticated - but this adds complexity ...and complex things/plans breakdown and require more resources and more people. More people means adding people with doubts, and the chances of leaking. Plus more resources, which brings attention to the operation. And as you add more people and resources, the "downside" to being caught gets bigger, so you try to reduce that risk by making it even more "foolproof".
If you are one of the 12+ people supporting the operative, and you have a 50/50 chance of being caught and spending a very long and nasty session in jail - even before you get your day in court - and you have no chance of the "ultimate reward" .... don't you think you might start having doubts, and talking to people? Sometimes the wrong people?
I don't buy for a minute all of the stories of traffic cops stopping a car for a routine check and finding "bad things" that were going to be used. The intelligence services have, imho, a pretty good idea of what is happening in these groups, and use these innocent looking traffic stops (and other coincidental discoveries) so that their undercover agents aren't suspected.
That is the value, imo, of the security checks. The barriers are are high enough to get the "bad" operations big and cumbersome, and to make the plans too complex to escape notice by the authorities. It's the planning and organization of getting past the security checks that the authorities are looking for. Once that "bad thing" is in the airport, the authorities have already lost most of the game. Then the security screening is just a last ditch attempt to catch something.
The real danger is the single lone-wolf person with a grudge, who hasn't planned in advance, and doesn't really care if they get caught. They have a 50/50 chance of getting through because the only security layer at that point is the security checkpoint. The intelligence services will not have picked them up, nor will the no-fly list incidentally.
.... all of this is just mho, of course..... read the later john lecarre though, for more chilling details....