In responding to @John_Harshman, Kojonan writes:
I find this to be a problematic argument. A reductio ad absurdum relies on showing that the opposite scenario leads to absurdity or contradiction. This is not clear here.
Firstly, your claimed “absurdity” itself relies upon an absurd hypothetical.
Secondly, your hypothetical does not in fact rule out non-divine explanations, such as a hallucination, or a very high-tech trickster – so it fails to demonstrate that that, even if we accept your absurd hypothetical, the result is an absurdity.
Finally, even if we accept it is an absurdity, this would only require a minor alteration to the contention:
… to read:
God explains anything equally well and equally poorly, except for something that only the existence of God could possibly explain (acknowledging that it may be impossible to specify such a thing, even hypothetically).
It is however unclear that this caveat has any real world implications, so it can probably be omitted for the purpose of brevity.
Addendum:
On further reflection, it occurs to me that a tighter formulation might be:
An explanation that explains anything equally well is inferior to any other possible explanation. (With the corollary that the explains-anything explanation should only be accepted if all other explanations are irrefutably impossible.)
This of course leaves out the possibility that you have more than one explains-everything explanation. Reformulating for this case:
All explanations that explains anything equally well is inferior to any other possible explanation. (However, if all other explanations are impossible, we are now in a quandary as to which explains-anything explanation to accept.)
I would hasten to add that both of these formulations are equivalent to @John_Harshman’s in real life situations, the reformulation is merely to nail down avoidance of the reductio ad absurdum accusation.
Further Addendum:
On still further reflection John’s formulation does not in fact fall into the reductio ad absurdum trap. He does not state that an explanation that “explains anything equally well” is an invalid explanation, merely a ‘poor’ explanation – and you would still accept a poor explanation, by default, if it was the only possible explanation you have left. At best my reformulations merely make this (perhaps) clearer.