First the easy part. JPEG Normal vs Fine/Superfine.
The only difference is how much information is discarded during JPEG compression. Is not so easy to see the difference at 100% magnification in a monitor from the original image, but if you enlarge it at 200 or 300%, or apply some sort of sharpening, the differences are there. Also the noise level is more apparent in Normal, and if you level postprocess the picture, to recover shadow detail for example, the compression artifacts begin to show their ugly face very quickly. Is also critical if you plan to make large prints. Always, the best method to evaluate it is printing the picture and evaluating it in paper, not in screen. If your final target is the web, there is no difference, because you will have to resize heavily the image (to 3:1 or more) so any gain in definition is lost during the resizing.
Now the hard part. JPEG vs RAW
I think it will be easy with an analogy. RAW is like your film negative, and JPEG is like your photo print. Every digicam has a monochromatic CCD as the image capture device, with a RGB colored filter above it to capture the color information, so 50% of the pixels capture the green component, 25% captures the blue and the last 25% the red one. But the image is essentialy a 12-bit greyscale one. To generate the JPEG color image from this raw data (hint here) the camera interpolates the missing color pixel for each RGB component, applies the white balance, reduce the color depth from 12 to 8 bits, and the 4:2:2 color space subsampling (after applying the color saturation and contrast selected by the user), resized it if needed (if the user chooses a smaller resolution) applies the sharpening filter, compress to JPEG, and stores it in the flash memory card. The RAW format makes none of the above. It's the 12-bit mono raw CCD/CMOS umprocesed data stored in the flash memory card, only with a lossless compression applied, and all the camera settings (white balance, sharpening, contrast, color saturation) stored as info. So, you can make all the color, contrast, etc. selections in your PC after the photo taking. If the camera had choosed the wrong WB, no problem, just select the right one. If the picture is too or too little contrasted, change the setting, and then store it as a 16-bit TIFF. You will get a non-compressed, 16-bit (or to be fair, 12-bit), 4:4:4 color subsampled image, much better than the JPEG one for postprocessing. More depth, better color resolution (in the reds and blues) and no compression artifacts. Also the RAW file is read-only in your PC, so you can't accidentally overwrite it as it can happend with the JPEG files.
Also RAW files has 12-bit in a linear gamma (gamma 1.0). JPEG or standart TIFF files have a non-lineal gamma (usually gamma 2.2) so the 12-bit image is gamma corrected during conversion, and, if saved as a JPEG or 8-bit TIFF, a lot of intermediate values, mostly in the middle to bright values range, are discarded. When the image is saved as a 16-bit non-linear TIFF, no values are discarded, because the 16-bit dataspace "extra room".
So, the RAW is your negative, as is the exact image captured, and the JPEG is a copy, because the camera has already applied a lot of transformations and cropping even before the picture is stored in the flash memory. You need more work and time to process a RAW image, but, at least for me, there is no doubt which format to use.
There is an article in luminous-landscape.com, titled "Expose (to the) right" where the value distribution is explained, and some tips to better exploit the linear space are shown.
The article URL is:
http://luminous-landscape.com/tutorials/expose-right.shtml