Why spoil the analysis of a beautiful data set by saying the wrong statement
"The Universe is flat"
when the correct statement is that
"the *observable* Universe is *approximately* flat"
? The WMAP data do not change this situation, though 1sigma = +- 0.02 gives 3 sigma +- 0.06 which means that the curvature radius is at least 12 h^-1 Gpc instead of 9.5h^-1Gpc (for +- 0.1) in comoving units at 3sigma. It's still like saying the Earth is flat because Africa is flat. Well, the Earth *is* (nearly) flat.
This is, indeed, further discussed in "WMAP results" thread on sci.astro.research. (Some postings are lengthy so - at least for this reason - I'm not going to re-post them here - just read sci.astro.research yourself. ;-)
Andrzej