We must always remember how rare terrorists are, and what that means for
any test you apply to a person.
You have a test that flags terrorists 99% of the time, and also flags innocent people 1% of the time. The test flags a person. What are the odds that this person is a terrorist?
Answer: roughly one in a million, depending exactly on the terrorist:non-terrorist ratio in your sample population. Since there are millions of normal people for each terrorist, a small false positive rate still grabs so many people as to make the test nearly useless.
And of course it goes well beyond this. Since actual terrorists are so rare, you can never confidently state that a test will flag terrorists 99% of the time. You can easily measure the false positive rate, but it's basically impossible to measure the true positive rate with any accuracy. You can guess and extrapolate, but when it comes to actually measuring performance in the real world, you'll have a very difficult time distinguishing between a test that flags 99% of terrorists, and a test that flags 1% of terrorists.