[ru999gol]

extraordinary claims require extraordinary evidence.

A magical, invisible, omnipresent, omnipotent "all known laws of nature defying"-being that answers prayers, requires significantly more evidence than a small effect in some meta study.

[pdpi]

Absolutely. My point is that "I'm an atheist so I'll take the option that agrees with my personal beliefs" isn't a valid line of reasoning. You could symmetrically argue "I'm a catholic so I'll take the option that suggests there is an effect, albeit a small one". Now we're just arguing whose personal beliefs are right.

A more properly scientific line of reasoning would be to instead say that you'll assume the small-to-nonexistent effect is nonexistent until somebody produces a model capable of predicting the proposed small effect, or a myriad other arguments along the same general lines.