Ah, it's pretty much as I had thought -- we don't have any direct evidence that these planets (Kepler-62 e and f) are water worlds, but they could be. Sasselov's argumentation was that they are more likely to be water worlds than rocky based off of trends we've seen with the observed densities of similar planets. (Which is plausible, but I'd like to see more data.) The rest of his paper has some really interesting discussion regarding the various types of water-planets that can exist, their atmospheric properties, and how we can characterize them in the future with spectroscopy.
But as far as Kepler-62 goes, the best we can do right now is say that there are at least five planets, two of which are super-earth's in the habitable zone, and we don't know what their bulk composition is. To know that we'd need to either get radial velocity measurements, or use spectroscopy. Unfortunately it would be exceedingly difficult to accomplish either of those for this system.
