First, let me state that I am far from convinced about this claim. What do we really know about the planet? We know its mass, which we have measured by the planet's gravitational pull on its parent star. We also know its diameter, because the planet passes in front of its star as we see it from Earth, and we can measure how much light it blocks out. Because we know its diameter and its mass, we also know its average density. In these ways, the planet is very close to a twin of Neptune.
But this is really all we know for certain. We can guess how hot GJ 436b must be based on how close it is to its parent star (this works well in our solar system for Mercury and Mars, it's close for Earth, and fails miserably for Venus). But we have no direct data on what the planet is made out of. The claim that the planet must be made out of "hot water ice" is a guess. An educated guess, but just a guess. So don't start buying beachfront real estate on GJ 436b yet.
Why would astronomers guess that the planet is made of "hot ice?" First, in our solar system, the planets Uranus and Neptune, the planets closest in size and average density to GJ 436b, are thought to be made of ices (including water ice) covered by an extraordinarily thick atmosphere of methane. So, it is not unreasonable to guess that all planets the size and density of Neptune and Uranus look like Neptune and Uranus. But, again, this is just an educated guess.
How can ice exist if the planet is hot? Look at the phase diagram for water below, which shows what form water will take as a function of its temperature and its pressure. For temperatures and pressures on the Earth (about 250-310 degrees Kelvin and a pressure of about 100,000 Pascals), water is near the point where all three phases of water -- gas (tan), liquid (green), and ice/solid (blue) can exist. On the diagram, notice that even at extremely hot temperatures, like 600 Kelvin (by chance, about 600 degrees Fahrenheit), water can be in the form of ice if it is at high pressures -- at least 10 million times the pressure on Earth's surface.
So, if GJ 436b has a very thick atmosphere, like Uranus and Neptune, the pressure inside the planet can be high enough to make water turn into ice, even though it would be very hot! This ice is not quite like ice on Earth -- the structure of ice crystals on Earth is not the same structure of ice crystals at high pressure. And, if this hypothesis is right, most likely the atmosphere of the planet would go through a gradual transition from air to liquid to solid. This is quite unlike the Earth, where there is an abrupt transition from the atmosphere to the sea and again from the sea to the sea floor. On GJ 436b, there is probably no real "surface" to the planet.
So, as I've said before, take this new press announcement with a grain of salt. Until we have observations showing us what a planet around another star is made out of, we can only make educated guesses. And the makeup of this "snowball in hell," GJ 436b, remains just that -- an educated guess.