I think nil.to_s == "nil" is pretty reasonable. At least, it gives .to_s an elegant definition that does not need to case-analyze the nil-ness of the elements of an array.

In any case, this just reinforces my point that Ruby emphasizes ease of use, not simplicity.