Even if we attribute all warming over the past century to man made greenhouse gases (which we have no basis for doing), the observed warming is only about 1/3-1/6 of what models project.
His point is that the models generally tend to predict more global warming than what we have observed, so that they are inherently alarmist.
My view is not so much that the models are biased in one direction or another (I cannot claim any expertise in that area), but that there is simply not enough data relative to the complexity of the process to have any confidence in a model. I would not be surprised if, as Lindzen implies, global warming proceeds more slowly than the models project. However, I also would not be surprised if the errors are in the opposite direction.