I have a calibrated wide-gamut monitor (Dell UP2414Q) and I use it with Windows 10 and Lightroom 6.
I wanted to print a photo with a Epson R3000 not installed locally, so I copied the ICC profiles of the Epson printer driver and I installed them in my computer.
This is the image. I exported it with AdobeRGB color profile, it may look slightly wrong or dull on normal monitors.
In Lightroom 6 I activated soft proofing, I selected the color profile of the paper I intended to use (I chose "Epson Photo Paper glossy" even if I'm using 3rd party glossy) and I saw that a big part of the image, the part that has very saturated colors (the whole liquid content of the glass, except for the floating lemon), was marked as out-of-gamut (destination gamut, the image fits my monitor gamut).
So I reduced saturation until the image fit in the destination gamut (I found out later that I could have used highlights with similar results). Well, the image looked dull. A lot.
Then I exported the modified photo and I told LR6 to produce a 16 bit TIFF with the gamut of the ICC profile I used for the soft proofing. I also took the unmodified, still saturated photo and I exported in the same way, but using AdobeRGB as color profile.
I printed both, same page, side by side (two different prints, I rotated the paper in between). I followed the instructions by Epson and I opened the photos in Photoshop, I told Photoshop to work with the embedded color profile, I told the printer driver to use "Epson Photo Paper Glossy", and I told the printer driver not to use any color management, to let Photoshop do the work. The instructions are here: http://files.support.epson.com/docid/cpd3/cpd39134.pdf
When I collected the prints, I noticed that... in both cases they looked as I saw them in the monitor just before exporting: one dull as I saw it after reducing the saturation, and the other, the unmodified one, basically just like it was when I got the gamut warning. Saturated. Well, just a slight bit less, it's difficult to compare paper and print exactly.
So it seems the Epson ICC profile was either too narrow (it expected a much smaller gamut than it was able to get) or I made a mistake.
To cross-check, I downloaded the ICC profiles from Saal-Digital http://www.saal-digital.de/service/icc-profil/ (SaalDigital_SoftProof_Fuji, Poster_Flex, Poster_Metallic) and they all show marked regions outside destination gamut about as big as the ones of the Epson Photo Paper. Just like with the Epson Photo Glossy, getting those regions back within gamut means operating on saturation and highlights a lot and the image is no more punchy as in the beginning (even if on paper it CAN be punchy, as I tested myself).
Last test: I loaded another ICC profile from Epson, called "R3000 Standard". It shows much smaller out-of-gamut regions and getting them back within gamut is easy. The result is still punchy and it is basically what I see on paper.
However, I cannot choose every time the profile that gives a posteriori the best results... it takes away the usefulness of the soft-proofing process that is meant to be performed a priori, in advance.
Also that same image printed on "Flex paper" by Saal-Digital is punchy even if their own profile now tells me the image wouldn't fit the gamut unless I heavily reduce highlights or saturation.
So am I doing something wrong? what? or is there an issue with the profiles and Lightroom?