A repeat of this experiment, with a few thousand participants, conducted under strict control would probably settle that question. One can't help wonder, why hasn't this been done?

I guess ethical and practical considerations aside (you'd have to find participants that agreed to the experiment, as well as having no grieving and potentially experiment-disturbing kin by the bedside): conducting such experiments would not earn you any positive citations from other scientists who largely try to steer clear of anything that smacks of spirits.

Plus, what if Duncan MacDougall's claim is proven correct?