The probability of observer metameric failure increases as the r,g,b emitters become more purely centered on a single wavelength each.
The CIE research, even in 1931, however was based on mercury vapor lamps and monochromator to setup the color matching functions. So beleive it or not, narrow band RGB has been part of the research from the very beginning.
Some of our common tools are challenged by new technology. Colorimeters in particular, have difficulty measuring color temperature as the light sources become more centered on single wavelengths. Other than laser, none of the consumer sources have quite hit that yet. QD's are probably the next closest thing to laser at this point.
All the wide CCFL, RGB LED and most OLED can be corrected with a spectral data file. At least that's what the math says.
But the OP's question wasn't that hard to answer. A photo viewing booth with D65 lamps, or a well-known display type, like a CCFL or white LED lit LCD, or a plasma display, aligned to D65 with a suitable instrument, is plenty good.