It is impossible to judge sharpness and fine detail contrast from a web size print. If you want to discuss sharpness, please post raw images or at least 1:1 crops.
Personally, I am a landscape shooter. Also, I just have MFD (P45+) for something like 3 months. The main difference I see is resolution, my DSLR is Sony Alpha 99 and the MFDB is a P45+, the resolution difference is quite visible on screen, but I guess not very visible in prints (OK, I see no difference in A2 prints and little in 70x100 cm (34x40") prints). I cannot say if any image I shot is from Sony Alpha or Phase One P45+, unless doing an analytical comparison or reading EXIF data. But, P45+ images are more yellowish and a bit sharper. They are also a bit noisier in the shadows. So, there are clues to pick up...
As I am a lanscape shooter, I am not very interested in skin tones. The cloth in the image is probably difficult, violet/purple are not spectral colors, they are a perceived mixture of two spectral colors. A violet flower can come out blue with one profile and reddish violet on another profile. Perception plays a major role. Accurate is not the same as pleasant. This has also a lot to do with contrast ranges. A good sensor has probably a contrast range of 13 EV, something like 1:8196 contrast ratio, when we show it on screen we need to compress it to around 1:500 (9 bits or 9 EV). In print we can only have something like 1:128 ratio, around 7EV. So tonal range is always mapped down. A straight mapping would yield boring images, so mapping usually involves pushing local contrast and saturation.
What is happening is that you can observe a few millions of colors. Those millions of colors are mapped down two three distinct intensities, R, G and B. That is all the sensor does!
The three signals than processed adjusted for white balance and converted to RGB by a color profile. You make a lot of adjustment to the RGB values. Those RGB values are put trogh some numerical processing to convert to screen or printer color space.
So it is a long process. A mix of spectral colors-> a single RGB signal -> Convert to input color space -> color space conversion for viewing -> adjustments ->color space conversion for printing. The sensor is only involved in the first step, converting millions of colors into three different discrete signals. Have you considered this?
very interesting insight about sensor technology. Thanks for that!
Regarding D800 files in Capture One, I have indeed tried to process the D800 files in it. I get a wee bit more resolution than in CNX2/ LR4, but the color rendering is totally off; no matter what color profile I use. To make matters worse, C1 doesn't have any option by which I can use a color checker to create a custom profile, Like I can with LR. In contrast, every Leaf file that I've dropped into C1 just looks amazing out of the box.
The below image is the best I can get out of the D800 after custom profiles, using a super sharp lens (The 85 f/1.8 G) and some high pass sharpening.
To my eyes, it still looks not quite "There". Maybe it's just me over analyzing.