If we adopt a flawed "evolution as agent" model, there is zero evidence that it has tried everything, or even most things, in order to optimize species.
If we adopt a more accurate model, evolution doesn't optimize or produce optimized systems. It can (but does not always) produce reliable (as in reproduction/survival) systems. There are plenty of ways Klotho could be beneficial but not produced by selection: mutations which increase production of this chemical may have occurred but had negative effects on fitness, or may have happened to occur along with other mutations with negative effects; there may not be any mutations which increased production (it's tempting to think of e.g. hormone levels as "knobs" for which there are mutations that turn up/down output, but that metaphor completely fails in many cases); beneficial mutations to increase Klotho production may be possible but not have occurred in nature because of chance or complexity of changes required to actually manifest the "right" change.