That sounds right. Maybe the the scientists are wrong. Water isn't H2O, its actually some other substance (lets abbreviate its chemical make up as "XYZ"). Under such a supposition, any substance that is not XYZ does not count as water -- this includes very watery collections of H2O. A moral here is this: in order to make counterfactual assessments one must make antecedent assumptions about actuality. For a discussion of cases like this and the even trickier case of Dry Earth see Korman(2006).