>>42169503/?
While this is great for comparing 2 spots on a background, most colors don't exist in that way. Our brain has all these tricks to adjust for bright and dim scenes with varying levels of contrast. Keep in mind, almost all color management today is done at "level 1", meaning each pixel on a monitor or camera, or raster cell on a printer is treated as a color in a vacuum. While this does produce pretty good results, its not the whole story.
The next level is color appearance. This describes perception that affects colors when their next to other colors, such as chromatic adaption, opponent colors appear more saturated when next to each other, simultaneous contrast, and other more complex stuff.
The main thing a color appearance model does is describe contrast and tonality as well as relative colors like greys and browns. The most common and best for most applications is CIELAB, which takes standard observer coordinates and mathematically transforms them into L* or lightness, a or green-red axis, and b or yellow-blue axis. The distance or deltaE between different colors is roughly how different colors look, but generally CIELAB underestimates color differences between low saturation colors greys and pale colors, and overestimates differences between saturated colors.
