Using Chesterton's Fence here is a little misleading.

The whole rationale behind Chesterton's Fence is that clearly someone put the fence there, and it seems pretty likely that whoever that was was just as capable as I am of concluding (given what I know) that putting a fence here is absurd, and it seems pretty likely that they know everything I know, and therefore I can conclude with reasonable confidence that they knew relevant things I don't know that made them conclude that putting a fence here is worth doing, and therefore I should significantly reduce my confidence that putting a fence here is absurd.

Using the same rationale for natural phenomena doesn't really work... there's a reason it isn;t Chesterton's Fallen Tree.

You can, of course, put natural selection in the role of fence-builder, which seems to be what you're doing. But actually there's lots of areas where humans are smarter than evolution. At the very least, humans respond to novel situations a whole lot faster.