Economists are currently running an experiment that dwarfs the LHC. Well, we're at least observing as the effects of stimulus to play out over the world. Of course, when the LHC gets running, a lot of physicists will update their beliefs about the nature and structure of the world. Five years from now, economists who believed large stimulus was an appropriate response will still mostly believe the same, and the economists who advocated no/little stimulus will still think it's the right move. But doesn't this prevent us from calling economics a science?
Not really. 'Is fiscal policy effective?' is a hypothesis in the same ballpark as 'Is the Standard Model right?', and if I understand correctly that question can't be completely answered by the LHC either, and we don't have anywhere near the controls the physicists do.
I concede we're making progress. For example, I think we at least generally agree now as a profession that there's a fairly significant welfare cost to business cycles because of heterogeneity, but don't hold me too strongly to this - I'm just coming off the tail end of H1N1.
POSTSCRIPT: At least I resisted the temptation to write a 'state of macro' post.