Evolution is optimizing for survival, not crop yield.

That will only be the case if light absorbence is the primary limiting factor for the plant. If something else is a bigger limiting factor (e.g., nutrients, pests, water, temperature, pollinators, and so on), then a more darkly colored plant will not be meaningfully more likely to survive and reproduce than its lightly colored neighbors.

Outside of human-directed breeding, we typically only see darkly colored plants in locations with low light and with few other evolutionary pressures: deep understory tropical plants, for example.

Meanwhile, in locations with high light and other evolutionary pressures, plants can sometimes evolve to be quite pale. For example, desert succulents are often very light colors because they need to deflect light to keep cool and preserve water.

I honestly doubt it’s in a global optimum, as evolution is easily stuck in decent local minima.

Simple example: “C4 carbon fixation”, which allows photosynthesis to proceed with much less water available, only evolved ~30 million years ago. This was the revolution that allowed modern grasses to emerge, and thus changed enormous barren deserts into the savannahs that early humans made their home. Despite being such an improvement, this is not yet a “standard”; trees, for example, still use the “old” inefficient “C3 carbon fixation”.

Despite photosynthesis having been around for hundreds of millions of years, fundamental improvements still happened just 30 million years ago. I would suspect that there is more that can still be improved.