CCDs are much better ...
Once again you are comparing sensors that different in multiple respects:
- CCD vs active pixel CMOS
- larger vs smaller sensor
- different CFA designs, likely aimed at different priorities of the target users (35mm and MF) as Doug Peterson has nicely discussed in his recent article at this site:
Read ...and sometimes
- different lenses, shutter speeds, effective aperture sizes (affecting out-of-focus and diffraction effects)
Strangely, the same comparison have on occasion been used (even by the same people!) to argue for different conclusions:
- CCD vs CMOS is the sole reason for the difference.
- Format size is sole reason for the difference.
With respect to "usable dynamic range", there are simple physical reasons why a larger sensor with slightly lower engineering dynamic range can still have better handling of shadows at all but extremely deep shadow regions: in moderately deep shadows at low ISO seeds (say two to four stops below mid-tones, maybe five to eights stops below full well exposure) the signal-to-noise ratio is dominated by photon shot noise rather than sensor dark noise, and then the larger sensor has a natural advantage through gathering more photons, whereas engineering DR only looks at sensor dark noise, not shot noise.