Ren mentioned another model of light, which makes me wonder what models are out there and why they are needed. Is the linear model not sufficient?
The superposition example here is of blue and yellow lights; we can see that the combination of both blue and yellow lights results in the distributions being directly "added" together, and so we get a superposition of SPDs.
This superposition example makes sense since each color has it's own unique spectral power distribution. Given multiple color sources, measuring the spectral power distribution of all of them combined makes sense to superimpose the individual power distributions on top of one another.