Kudos to Apple for making the new iMacs wide-gamut. But I don't know that it's too common in the consumer space yet, mostly because once you start using multiple color spaces, you run into the problem of images created in one color space being shown in another. If the image data isn't tagged correctly, things will look over/under-saturated on different monitors. (E.g., is rgb(255, 0, 0) the reddest red in sRGB or the reddest red in P3/Adobe RGB?)

I suspect that when monitor manufacturers tire of competing on pixel density, brightness, 3D, and size, they'll eventually get around to competing on gamut. Maybe Apple will lead the way. A lot of photo/video cameras already capture a larger gamut, so monitors are the last piece of the puzzle.

Unfortunately as the gamut gets larger, so does the visual difference between adjacent colors. You need more than 8 bits per channel from end to end if you're going to use anything wider than sRGB.

Any RGB based color space will be triangular shaped; if it's to be physically realized in a display, all 3 vertices must lie within the CIE gamut. Since the CIE gamut isn't perfectly triangular you will always be missing some colors outside of the triangle no matter what you do. You could do much better by adding a 4th primary to create a trapezoid, which would more completely fill the space. As the entire industry is RGB based this would be completely impractical, but it would be fun to see as a demo.

Is the CIE gamut "it"? In other words, are all colors perceivable by a typical human represented by it? I briefly scanned the Wikipedia article but wasn't any wiser.

Yes, the CIE 1931 color space is generally accepted as the limit of human perception. I don't know how much variability there is from person to person, since it's based on averages. There are a small number of Tetrachromats that can distinguish a greater variety of colors, but I believe those colors are all still within the gamut.