I've tried the soft-proofing thing and didn't find much value in it. Nothing posted above has changed my opinion.
With all due respect, I might suggest that if soft proofing hasn't helped you in the past, that possibly, you haven't actually learned how to soft proof?
When one soft proofs using both the output profile _AND_ checking the options for the Display Options (paper white and ink black) it will of course, make your image look like crap. That's what it's designed for in Photoshop. To simulate both the final gamut of color as well as the dynamic range of the print (the contrast range between max black and paper white).
In my experience, assuming an accurate display profile and output profile, Photoshop's soft proofing is about 90% accurate...the limit is that fact that most displays these days can't show all of the gamut of color that printers such as the 11880 can print. But a lot of those deep, saturate colors don't often appear in "natural" photographic images any way.
The other disconnect is often the fact that people get used to seeing the RGB image on their display and when turing on the "make it look like crap button" (soft proofing) they may not be evaluating the results properly. First off, the eye does what's called "white adaptation" which means the eye looks at the lightest tone value in the field of vision and assumes that to be "white". Therefore, if you have a lot of white UI showing (the palettes in Photoshop), the soft proofed image will appear darker and flatter than it should based upon the white surround. Therefore, it's important, when soft proofing, to turn that Photoshop UI off. Hit the Tab key to hide the UI and the F key to give a dark gray or black background.
Then the problem is that often, the viewing conditions of the print may not be consistent...i use a GTI viewing booth (set 90 degrees to the display) that allows me a digital dimmer) so that the luminance value of the display and the paper white of the print match.
Done properly, the resulting soft proof display view and final printer piece allow a very accurate comparison. About 90% accurate in my experience. Which is close enough to adjust an image accurately for the final print without actually making the print.
There are also issue using actual printed downsized proofs to predict the final large size print. JP Caponigro says that large prints tend to need a tone/contrast adjustment over the same image printed small. Large prints need to be darkened a bit from a small image proof. He actually has suggestions regarding the amount to alter the image based upon the square footage if the small print vs the large print. Check his web site at www.johnpaulcaponigro.com/
in the Techniques section and look for a PDF named tech proof for scale.
If you are printing on a watercolor paper that has a substantial texture, the small print proof will also have problems predicting the extent to which the texture will alter the ability to predict the impact of texture on the details of your image. JP also suggests printing smaller swatches of the image at the final large print size on the same paper you'll be using for the large print. This allows you to fine tune the sharpening for the final detail in the print.
All of these factors impact how one arrives at a printable image for the final print. If all factors are taken into account, the final large print should not be a surprise.