I suspect there might well be an evolutionarily balanced tradeoff between "construction cost" and "lifetime energy output" of a plant's photosynthesis machinery - as well as a likely overabundance of available solar energy in terms of what a plant actually can use. (Why build a potentially more complex and fragile photosynthesis pathway that's 50% more efficient, if you can instead just grow twice as much leaf surface area?)
But there are places where plants compete fiercely for small amounts of sunlight, such as in forests (especially tropical rainforests). Doesn't your theory suggest they'd use more efficient systems there?
Don't forget plants get the energy back from making a leaf in a few weeks. Solar panels take considerably longer. Generally plants are optimized to compete with others in their niche and as such it's exponential growth that's most important not total energy capture.
Is it actually tropical rainforests, or temperate, that limit sunlight most? Old-growth PNW forests often have large patches of western hemlocks that produce so much shade that no other tree can grow under it, not even conifers - and unless there's a fire to clear some space out, it can stay that way for centuries.
Selection pressure may well have shifted towards C4 in the future if we hadn't come along and been kind enough to dig up some fossilized carbon and put it back into circulation.
