Cosmic deceiver?

My answer to a question:

1. We can't extrapolate from examples of God deceiving the wicked in Scripture because they carry no presumption that God deceives humans in general. Those passages are confined to a subset of humans, and not a random sample, but humans who are punished for impiety. So not every human being meets the necessary condition. As such, Scripture provides no justification for belief in universal divine deception.

2. Of course, even apart from such passages, it's possible to entertain the idea of universal divine deception. If Scripture never spoke to the issue of divine deception, even in that limited context, we could still toy with the idea. 

3. Seems to me that fallibilism is self-refuting. Some beliefs are only properly doubtful if other beliefs provide a benchmark. We find a claim dubious or unbelievable because it conflicts with a belief we find persuasive or compelling. Of course, we can still be mistaken, but I don't see how we can be universally mistaken, since erroneous beliefs presume some true beliefs as a point of contrast to provide a necessary standard of comparison. 

Put another way, how is it possible to coherently argue for fallibilism? You have to provide reasons. But if all the reasons for fallibilism are fallible reasons, isn't that self-referentially incoherent? 

4. Mind you, rejecting fallibilism tout court doesn't indicate where to draw the line. In principle, there's a spectrum ranging from mostly false beliefs to mostly true beliefs. 

5. A problem with variations on the Cartesian demon is why a being that's omniscient/omnipotent or nearly omniscient/omnipotent would take any interest and find any satisfaction in toying with and fooling creatures so vastly inferior to itself. Even if he's malevolent, what's the fun in outwitting creatures who are so incomparable below his own level of intelligence? What's the point of making creatures he can effortlessly outsmart? It's like a chess genius inventing a stupid chess computer to play with. He always beats the computer. It isn't hard. There's no challenge, no surprise, no risk of defeat. Even if the being is malicious, what does he get out of that exercise? How is that entertaining? 

6. There's also the question of whether good and evil are asymmetrical. When we hypothesize an evil godlike figure, evil is a comparative judgment. It presumes some concept of the good. It presumes an ideal from which the evil deviates.