Re: mass vs density
@D@v3: we're 70% _surface_ water, but by volume... if you assumed we average six km of water on the surface, and have about a 6000 km radius, that'd make us about 1/3 % water by volume.
@LeeE: as I described, those numbers indicate something denser than _lead_, not just for the core, but for the entire planet. It'd need a core of something really unusual, like mercury or gold or osmium or uranium -- and a whopping big core, too, and still need some heavy stuff around it.
The radius was determined by seeing how much the star's brightness dropped when the planet went in front of it (which wasn't much; the planet is a _lot_ smaller than the star, so it blocked barely enough light to be noticeable.) The mass was determined by measuring the radial velocity of the star relative to us: when the planet is moving away from us, the star is moving (very slightly, because it's a lot heavier) toward us, and when the planet moves toward us... you get the idea; measure the difference in speed between "moving toward" and "moving away", combine with an estimate of the star's mass and the distance between the star and planet, and you can get a mass for the planet.
These are all fussy measurements we couldn't even have done a couple of decades ago, and we're still at the point where the errors are a good percentage of the quantities being measured. The most likely scenario is just that the radius was underestimated by, say, 10% (which would drop the density by 30%) and/or the mass was overestimated by a bit.
Standard practice in a scientific paper would have been to say something along the lines of "the radius is 1.4 +/- 0.2 times that of the earth, with a mass 7.0 +/- 2.1 times that of the earth." The sigmas probably got dropped in the press release.