You are correct that, beyond a point, the effects are substantially nonlinear. Of course, our reactions would also not be static.
My point was not prediction of specific outcomes, but rather to illustrate the tremendous disparity between the two scales - even in a year when we lost an order of magnitude more people to terrorism than is typical.
(You are also correct that there are substantial effects beyond the direct loss of life - but that is true with heart disease as well.)
"The entire country would be in ruins. There would be no more cities"
This is still exaggerating the scale.
If we say 3000 9/11 sized-attacks and 3000 dead in 9/11, that's 9 million people. In that time, our population grew by 50 million.
If we say the 9/11 attacks destroyed 10 million square feet of office space, that's 30 billion - about twice what we added between 2003 and 2012 (per http://www.eia.gov/consumption/commercial/reports/2012/build...), so plausibly about what we've added total in that time.
There may be "no more cities", but it would be due to a response of radical decentralization (probably not a bad move, in that situation) not because the cities were physically destroyed.