The basic underlying reason, I believe is selectivity.
When fishing during selective feeding, one may catch some fish with a spinner that does not match the color. One may then think that matching the color is not necessary. But all one has really proved is that some
fish are not selective to color.
Fish behavior is population based. They do not all behave identically and for each behavior we can graph the fishes behavior and the population distribution will lie under a bell curve. For selective behavior, as the population become more and more selective, the curve will shift to the right and the mean behavior will become more selective. As the behavior becomes more clustered, the distribution of the curve tightens and the slope of the curve becomes steeper and the peak taller.
The diagrams below are used to illustrate what I think happens during the development of selectivity. They are not
meant to state that all populations can be fit into the symmetrical perfect bell curves below.
The Venn diagram below demonstrates what happens to a population of fish during hatch as the population moves form nonselective feeding to selective feeding.
The same concept can be shown as a bell curve population distribution with the horizontal X axis representing increasing selectivity as one moves to the right. So the pink bell curve population is more selective than the blue bell curve population.
If the hatch lasts long enough and is heavy enough, the population distribution can cluster and selectivity criteria can become extremely strict. Again this is represented below by a move from blue to pink.
In the graph below during a heavy hatch the fish may show the wide blue distribution of day 1, then day 5 they demonstrate more consistent search criteria and on day 10, their criteria become much more strict.
There are places where fish, especially the largest fish are extremely selective. They will be selective to color.
So you may not need to carry those colors but this does not mean that a need for those colors does not exist.