I think the mistake you are making is switching from a strictly logical basis to how we evaluate empirical claims.
It’s easy to show the flaw here with an argument from absurdity:
Let’s say that our world was like many people’s conception of hell: burning fires, skin healing just so it can be burned again, all that jazz. And heck, we can even throw in that God himself appears to smile and laugh at our suffering.
Now; according to Plantinga’s reasoning, we can still argue that, technically-speaking, none of that would prove that God is not omnibenelovent. There might be reasons like free will that make even hell the best possible world overall. And hypothetically there might be sound reasons for a caring God to appear like he relishes suffering. To which I say (inbetween screams) “sure…that’s implausible, but possible”.
But your claim is that none of that would even give us reason to doubt the existence of an omnibenevolent God. You are going much further than the logical claim, into absolute absurdity.
Heck, by this line of reasoning, all of science would be irrational. Because we don’t strictly prove things in science too, all we ever have are levels of confidence.