Michael Grunwald has written an article for Time arguing that the effects of the BP oil spill have not been as bad as initially feared.  There's a phase in disasters--often quite long--when people start treating the worst-case scenarios as if they are the most likely scenarios.  I suspect this is a bit of hard-wired evolutionary programming, and if you're a hunter-gatherer tribe at risk from disasters, this is probably quite useful. 

Let me be clear that I'm not saying "people overreact to environmental disasters!"  The initial forecasts of the dead from the World Trade Center were in the 10,000 range, which turned out to be about 4 times to high, and were treated by most of the people in New York, at least, as the actual numbers until body counts revealed that the number was quite a bit lower.  I think it's safe to say that we overreacted to that tragedy in quite a number of ways, for reasons I also think are probably hard-wired.

So it's not surprising that the actual effects we're seeing turn out to be not-so-bad as the projections.  But of course sometimes they really are that bad--Haiti seems to be going worse, in some ways, than we expected.  And Jonathan Adler sounds a very sensible note of caution about popping the champagne corks on the BP spill:

The reality is we have sufficiently little experience with this sort of thing, so we don't really know how bad an environmental disaster the spill is and will be -- and may not for some time.  Some of the easier to measure projections may have fizzled, but the spill could still be having significant as-yet-unseen ecological effects that we do not yet understand -- and may not for some time.  Still, the Grunwald article is a useful reminder that we don't really know enough to make categorical statements about the likely ecological effects of the spill.

And of course, even if the worst hasn't happened, the results of the spill have been quite horrible enough.