This is an obvious exploration/exploitation optimization strategy. The child doesn't have infinite time and has the same distribution for available time in both

This is an obvious exploration/exploitation optimization strategy. The child doesn't have infinite time and has the same distribution for available time in both cases. A larger estimated state space therefore dictates more aggressive exploration, which means faster switching and slower settling. In more intuitive terms, the child is trying to maximize their chance of getting to play with everything, and while they don't know how much time they have, they have no reason to assume it's enough, so they move faster to cover more toys. Whether you think that this adaptation is a reasonable one, or whether these are useful observations for a child to use to inform their expectations about the world, I'm not sure I can say. But this is the rational strategy under straightforward assumptions about the utility function that our adaptations approximate. Furthermore, I expect that I would observe identical behavior in adults - for example, consider your takings at a buffet as a function of the number of options and your familiarity with them - and in fact that a non-negligible part of HN often exercises this same strategy consciously and rationally.