Great literature is an instantiation of how some of the smartest humans have thought about the human condition, with a degree of exploration much wider than what the real world affords us. Which means it is a way to explore a much wider swathe of mind space (I call it the demandscape), bringing all the usual benefits of that exploration of unknown brings you.

Not easily quantifiable in the abstract or before you actually do it, but undeniably useful nevertheless.

It falls therefore into the group of activities that are definitely useful when done but without clarity as to how it is useful, maybe comparable therefore to liberal arts education or taking time to explore carrier options or writing online or ...

All of the modern ways of solution analysis, in terms of what the effect is of doing a particular thing, is extremely low dimensional.

Which basically means that for most high dimensional activities you have to actively step beyond your analytical framework in order to get benefit from them.

This was my thesis for writing the Strange Loop Canon commencement address, in the sense of most advice actually being useless. But it is not just a simple statement, it permeates into all activities that have broad-based benefits like reading great literature, poetry, philosophy, history, biography and more.

Luckily this also gives an answer as to why some types of art or activities are not particularly generative or useful, e.g., about how much of TV even if it is beautiful and brilliantly written is not the same as what you would consider great art.