He recommends not rejecting the hypothesis outright, but putting judgement on hold in lieu of evidence. If evidence is presented then be prepared to re-evaluate. He talks about present day tools and measurements as well and the need to re-evaluate in the future should future data be obtained.
I think the problem he’s addressing is when people use beliefs in lieu of evidence. Believing the election was rigged is fine, but without evidence you haven’t given Sagan or other evidence-based evaluators reason to believe your claim.