I'm sorry, but I have to smite you, my friend. If I could smite you ten times in a row, I'd do that, too. You don't seem to get that for the last several years, the ENTIRE time we've all been waiting for this camera, we all heard nothing but "LESS MP, BETTER ISO!! LESS MP, BETTER ISO!! LESS MP, BETTER ISO!!". Here we are, on the day the long-awaited 5D III is released, and everyone is BITCHING about the fact that we "only" got a "measly" single extra megapixel. Well DAMN, PPL?!? If you wanted uberpixels, why did you demand less mp and ask for better ISO?!?!
Simply put, the 5D III is a FANTASTIC camera from a specs standpoint. Canon listend to ALL of their users complaints, and fixed just about all of them, from what I can tell! Like the 1D X, it STOPPED focuing on megapixels, megapixels, megapixels, and STARTED focusing on WHAT PEOPLE FRIGGIN ASKED FOR!!!!!!!!! We just got a whopping TWO STOPS of NATIVE ISO improvement!!! The Nikon D4 didn't change native ISO one tiny bit, and neither did the D800!!
1. many people were asking for about 30MP, although many were not, not sure the break down, but probably between 70:30 to 35:65.
2. The thing is it does NOT have 2 stops native ISO improvement. Even Canon has said some to much of it is just in cam jpg NR fakery. Canon says the 1DX is 1 stop better than the 5D3 and the 5D3 is 2 stops better than the 5D2 but it is beyond the laws of physics for the 1DX to be 3 stops better than the 5D2 (certainly barring a change to a radically different technology). While we may get 1 full stop better than the 5D2, which would be very nice, we might only get 1/3-1/2 stop better in which case it might not be all that different from the D800 SNR after doing complete normalized comparison.
3. The D800 uses an Exmor sensor and they have been getting very near the theoretical limit for low ISO DR compared to high ISO DR. Some of them have had 13.8 stops. The 5D3 has 14bit files there is no room for it to be better than the best Sony/Nikon sensors so far. Canon has failed to mention any ISO 100 DR improvement. Canon apparently turned down tech prior to the 5D2 that would have let them match Exmor DR because they wanted to milk more money out of what they had. There is a solid chance that the 5D3 might have say 11.5 stops DR while the D800 13.8 at ISO 100.
So for all the talk about low MP and awesome IQ it is not at all impossible for the 5D3 to have not only 50% less MP but 2 stops worse ISO 100 DR and perhaps only 1/3 stop better SNR at high ISO for all we know.
I sure hope that is not the case. In the best case comparative scenario maybe it pulls of 1 stop better SNR at high ISO than the D800 and 1 stop better DR at ISO 100 (this latter part is very unlikely, since it would require a huge step up from Canon in terms of DR and a near 1 step drop back for the D800 from the D3x sensor). If so that certainly would be nice.

My guess is the 5D3 will have anywhere from 1 stop worse to almost the same ISO 100 DR and 1/2 stop better high ISO SNR.
It's all guessing though. By mid-April we may know all the answers.
At this point we really can't say which sensor will deliver better in any respect other than the D800 has more max detail and reach at lower ISOs.
They did give us awesome AF though (on paper at least), which is, well, awesome.