From the article:
"Spergel and his team are now working with Weeks to see if they might somehow have missed the circles. And there is a further test of the dodecahedron model. It predicts that a key measure of the density of matter in the Universe, which governs its curvature, is equal to 1.013.
Completely flat space corresponds to 1, while values greater than or less than 1 would create a curved Universe. Observations of the microwave background radiation so far suggest the value lies somewhere between 1.00 and 1.04. Further observations by WMAP and other instruments should give a more accurate answer within the next few months."
From this, and other observations, I'll go with Weeks and not Spergel. I just don't buy the "six matching circles" requirement.
This has some implications for other cosmological models. If this finite model is correct, then several others are ruled. An eternal, inflationary universe is in the toilet, and the brane collision model also comes to mind.