But remember that the intensity of solar insolation would be weakest near the terminator, and you'd have the nearby glaciers acting as a thermal sink. The result should be a dry, cool environment. You wouldn't expect to see high surface temperatures anywhere except near the dayside pole. As far as topography goes, I think the article's model assumes that there isn't much vertical exaggeration because of the higher gravity. In any event, ice would just follow whatever regional topographic gradients already exist, and there would undoubtedly be nightside regions of tectonic uplift where massive continental glaciers simply flow around the rocky massifs. The key variable would be total available surface H2O: too little and all of it builds up permanently on the night side, flattens itself out, but none ever relaxes completely to the terminator. In such a model, oceans would be precluded.

It's a good point about global temperatures in a thick atmosphere, with Venus as an example. We don't know the actual atmospheric composition of Gliese 581c, so its atmosphere remains a free parameter in any model of that world's global geochemistry. However, it's important to note that Venus has a very heavy atmosphere, about 93 bars at the datum, and is mostly CO2. Also its cloud cover (the formation of which is still something of an unresolved issue) assists in heat retention, and most importantly Venus receives a great deal more insolation than would a planet orbiting Gliese 581, even at the proximity of planet c.

If Gliese 581c had an atmosphere very similar to Venus, it might also have a runaway greenhouse environment. If its atmosphere is substantially thinner, or contains a large partial pressure of infrared-inactive components (e.g. N2), a runaway greenhouse would be unlikely, as would a global distribution of intense heat. Gliese 581 is a very dim star; only about 1% as bright as the Sun. At 1 AU distance, the blackbody temperature of an Earth duplicate would be low enough to freeze out most of our atmosphere. Even as close as Gliese 581c is to its primary, its dayside pole would still receive only a fraction of the energy that the planet Mercury receives from the Sun.

About primordial H2, etc.: the gravity of Gliese 581c might make retention of even the lightest gases (H2, He) possible, and if the planet formed farther from its primary and wandered closer at a later date, this would be even more plausible. However, in our system the inner planets were stripped of their primordial atmospheres by intense solar winds during the T-Tauri phase of solar formation. A red dwarf would have a weaker T-Tauri phase, but would still exhibit a more energetic phase of solar activity as it formed. Gliese 581 is a variable star, which probably contributed even more violence to its birth. It's probable that inner worlds in the Gliese 581 system would have undergone some atmospheric stripping, making a modern H2 and He atmosphere around a terrestrial planet unlikely.

I look forward to future spectroscopic observations of Gliese 581c's atmosphere. Such data would settle many of these questions, and teach us a lot about planetary formation processes around red dwarf stars.