But I think he said what I said, and meant what I think he meant, he just said it better.
He is saying (to me) that there are holes in the theory (gaps in the record), and while he believes in the theory, owing to (as you and he see) "a preponderance of the evidence", he is at least willing to imagine observations or experiments that rely on those holes to potentially disprove the theory. I firmly do not believe he in any way meant to suggest that those observations or experiments would be falsified. If I get it right, his very Punk Eek theory uses those very holes as a sort of dis-evidence of the potential validity of his thesis.
(Didn't von Daniken find a radio in the Cretaceous? Or was it a Battery in Babylon?)
Not here he isn't.
He's simply talking about which theory has real information content and is subject to disproof by evidence that contradicts what the theory says. The actual information content of "Well, God could have done it that way" is very low. The statement works equally well against any conceivable "that way." It isn't testable, falsifiable, or scientific. That's what Gould is talking about.