Well what's happened is I'm reading all these forums about dynamic range, dynamic range, banding banding and it's messing with my mind man! LOL
Thats exactly why I'm trying to combat all that blather!
Don't listen to it...it has no bearing whatsoever on the cameras ability to take excellent photos...its just a bunch of tech heads who like to tear apart hardware and data at a low level and find its flaws (or lack thereof, if that tickles their fancy.)
Its all meaningless in the grand scheme of things! Learn how to use your camera, use it effectively and efficiently, and those issues will only affect a very small percentage of your total photos. If you need real evidence of whether any one of the cameras on the market today, including the 5D II and by extrapolation the 5D III, just look for photos created by them on sites like 500px.com or 1x.com. You'll never see any banding, fixed pattern noise, or DR issues.
To try and set your mind at ease a bit more, we've all been bickering about DR at an extremely low level, and in terms of extreme situations, where you might need to recover shadows by more than 4 stops. Unless you have rather unique photographic practices where you are regularly photographing scenes with unbelievable DR and can't use an GND filter to balance contrast, its highly unlikely you'll need to recover that much low-ISO shadow DR...ever. Using the more real-world DR numbers from DPR (as contrasted with the very low-level numbers from DXO that have been involved in most of the DR discussions on these forums lately), here is how cameras compare today, in stops of DR @ ISO 100:
|Sony Alpha A900||9.4|
|Canon 1D IV||8.6|
|Canon 1Ds III||8.6|
|Nikon D3s & D3X||8.5|
|Canon 5D II||8.4|
The grand total difference from the worst current Canon body, the 7D, and the best current Nikon body, the D7000, is less than a full stop. The difference is even less, from the 1D IV to the D7000, of about half a stop. In terms of more realistic, real-world shooting...you shouldn't concern yourself with the low-level nature of read noise and how it may affect the technical specification of DR.