Broken Window Fallacy?

Recent events in Japan have prompted discussions of the broken window fallacy. The typical chain of events goes something like this:

1. Natural disaster
2. Reporter (or, reporter citing an economist) notes GDP might increase because of disaster.
3. Economist (probably not the one cited by the reporter) yells "BROKEN WINDOW FALLACY!" And, presumably, scores points for being reasonable.

Today, I will venture into the land of heresy by suggesting those who point out that natural disasters might make people better off are not necessarily committing a fallacy.

A fallacy is a chain of reasoning whereby the premises do not support the conclusion. But consider the following chain of reasoning.

1. Individuals may have a bias whereby they do not upgrade appliances when doing so would make them better off by their own assessment.
2. Natural disasters force them to buy new appliances.
3. The net effect of natural disasters is ambiguous (i.e., it might be positive, negative, or zero)

Where's the fallacy? Note: I am not saying that natural disasters make us better off. I am saying they might make us better off if the behavioral problem identified is significant. I think it is unlikely that the benefits from upgrading our refrigerators would be so great that they would offset the buildings destroyed by the natural disaster. (Of course, if the underlying behavioral problem is significant, it might apply to more than just household appliances.) But this is an empirical question.

My suggestion: stop calling such claims a logical fallacy. Point out that an increase in measured GDP does not necessarily imply an increase in welfare and then move on. There is no need to paint those we disagree with as unreasonable, illogical, or stupid.