SHarpness is in the eye of the beholder. What looks sharp is a picture in which all the important objects are in focus. This means you need a big DOF if your subject is close to the camera ( within 10 feet, say.) As the sensor gets bigger the DOF reduces for a given aperture, because you have to move closer in to fill the frame. Or you have to put a longer lens on to pull the subject in to fill the frame.
If you compare two cameras each with the same FOV, but one has a 1/1.7" sensor and one an APS-C sensor, the APS-C image will look unsharp because the DOF is much less than the 1/1.7" sensor. On the 1/1.7" the actual FL might be 10mm, but on the APS-C it will be 35mm to give the same FOV, and the same size image on the display (assuming you set the resolution to produce the same size image.) Because the APS-C has a 35mm lens, the DOF would be much less than the 10mm lens on the 1/1.7" If you want the APS-C to look as sharp, you would have to mount a 10mm lens to get the same DOF. But you would then be looking at a very small image on the screen, and you would have to dolly in to pull in the subject to fill the frame. This situation is worse with FF sensors. Only if you shoot something far away (like a building) will the FF camera image look sharp, but things up close will be blurred. Whereas the 1/1.7" sensor will produce images with sharp objects in the distance and in the foreground.
This is why people who sell and market cameras show you pictures of charts, because they are so thin the poor DOF of the FF camera doesn't make the picture soft. So, camera testing labs can state high resolution using paper charts, without giving away the truth that in reality the camera has lower resolution a few millimeters in front and behind the chart.
Focussing accuracy with FF sensors is critical because DOF is poor with even WA lenses compared with 1/1.7" sensors. FF cameras have a hard time getting the focus right because DOF is so small. That is one advantage of the M cameras with mechanical focussing - assuming it matches the optics and sensor.
On LI, there are many out of focus images from FF cameras, because the DOF is so poor with even 50mm lenses. I see people struggling to get DOF using 15mm and wider lenses for nature scenes, which themselves introduce distortions and color issues, when anyone can make sharp scenics with a $300 small sensor camera.
Sharpness depends also on photo site spacing. The smaller the spacing the sharper the image because you also get more detail from closely spaced photo sites. That is how Fuji is able to get detail. By making the photo sites in a diagonal pattern, the vertical and horizontal spacing is smaller. And Fuji added an additional photo site in each grouping (super CCD HR sensor) so that there were five sites per pixel, not four as in the Bayer sensor. The sites are 2 microns apart in small sensor cameras, which gives better detail than the 4 -6 microns in FF cameras. To match a small sensor camera with 12MP, the FF camera needs 48MP.
FF cameras are good for product shots, and portraits. For everything else use a compact camera. It is very difficult to make pictures from an APS-C camera as detailed as a compact camera. I am currently trying to make my NEX-5 APS-C camera
make pictures as sharp as my 16 year old DMC-LC5 camera with dc vario SUmmicron lens. I have tested every lens with the NEX-5. I can get close in PP by tweaking sharpness and contrast, but I can't get the DOF even at f8, that I can get at f4 with the LC5.
SUre the noise is much lower with the NEX-5 at ISO 400 and up. But the Dynamic range falls off by 1 stop every time I double the ISO. And the NR in the camera smooths out the detail as ISO increases. It does make lovely portraits with the older lenses.
Things were different with film. The 35mm film had an equivalent resolution of a 128MP sensor on a FF camera (approx.) so detail was always good enough to project a slide to 10 feet wide (we did this routinely for commercial work.) It will be a long time to get that sort of resolution in our amateur equipment.
In summary: The maximum resolution spec of a camera is unimportant above 4MP, as anything that is more pixels than a monitor can show, has to be scrunched down by scaling the picture, and this reduces detail and sharpness, and you can't get it back in PP - all you can do with unsharp mask is make straight lines sharper and put haloes around tree limbs.
What you want for best detail is for the picture to be sent to the WEB without scaling. Right now the limit is 2MP (for browser viewing on LI), so shoot 2MP and fill the frame.
If you want a shallow DOF for portraits use the FF camera or move in real close with your compact camera - or step back and use a longer FL lens if you want to control the perspective.
If I had understood this years ago, I would have saved a ton of money by not getting suckered in to the "high resolution" cameras. I don't make prints any more, so I don't need more pixels to suit the printer (the printer needs more data to make the picture look sharp.) But if I did need to make a print I could scale up my 2MP image to 8MP in my editor, and the print would look fine - at least as sharp as my monitor looks, because a large print is viewed from further away. Here I sit 18" from a 24" screen and the pictures are sharp as a tack. If I printed a 19 x 12 picture from the 2MP/8MP upres image, it would look just as good from 4 or 5 feet away.
I once scaled an image to 8 feet x 6 feet and viewed the massive 150MB image on the monitor and the detail was good enough for the print (we never printed it.)
Remember that before 2002 or so, professional cameras had 6MP sensors, and picture books were made from them. Picture books aren't getting any better with the "high res" cameras. I have a lovely printed book of photographs of Maine and the photographer used a 5MP Digilux 2 for all of them.