And here's an argument that can be used wherever someone tries to pass off a merely possible (conceivable) explanation as the actual explanation for a phenomenon. I'll refer to their merely possible explanation as P.

1. The only reason in favor of adopting P is its mere conceivability.

2. It is rational to accept a merely conceivable explanation as accurate if and only if there is no other rival alternative explanation.

∴4. It is rational to accept P if and only if there is no rival alternative explanation. [From 1 and 2]

5. For any P, there is a rival explanation, which is also merely possible (conceivable): namely, we are part of a simulated reality and all phenomenon are mere appearances without any material casual relation to one another.

∴ It is not rational to accept P. [From 4 and 5]

You can run this argument with any rival alternative hypothesis, for example, that everything I experience is a dream.

Here are some general criticisms about the form of reasoning which you noted people use.

A) Mere conceivability is not sufficient grounds upon which to make a claim about actuality/reality. It is conceivable that ghosts materialize and dematerialize at whim, but we have no independent reason to suppose that actually happens. Mere conceivability doesn't imply plausibility.

B) This seems like an ad hoc modification of the "ghosts exist and cause phenomenon" hypothesis, which is motivated by a prejudiced commitment to the hypothesis, not independent evidence or reasoning.

C) Presumably, someone who accepts that hypothesis does so because, if it were true, it would explain some phenomenon—say, an object mysteriously turning up in a weird location. There are innumerable rival conceivable explanations that can be imagined with respect to any such phenomenon, some more plausible than others. The only rational way to select one out as the most likely explanation is by appeal to independent evidence or reasons.

Ah, right, that's a fun ad hoc explanation. I'll work on an argument addressing it! 😁