One HTML page: a person holding an umbrella in rising wind, with a canopy that flexes, inverts and gets torn sideways, and rain that follows the wind. Everyone in the audience has held an umbrella in a gale, so wrong physics is obvious without any measurement.
No instrument needed. The viewer's own memory of an umbrella is the answer key, which is why this one works on camera.
Drive the wind up and back down. Inversion is the moment to watch: it should be a state the canopy enters under load and can come out of, not a sprite swap at a threshold. Ask whether the umbrella fights back on the way down.
Does the rain agree with the wind? Drops falling straight down while the canopy is horizontal is the single most visible failure and takes one second to spot.
Does the person do anything? Lean, grip, footing, bracing were all named. A figure standing bolt upright with a thrashing umbrella attached has skipped the coupling that makes the scene read.
Check the wind readout against what is on screen. A number climbing while nothing changes is the same class of failure as an FPS counter reading 60 at four frames a second.
The last section asks the model to name what its own umbrella gets wrong. Compare that list against what you saw. The gap between the two is the honesty score, and it is worth more here than the prettiest canopy in the round.
You build physics toys, and you have held an umbrella in a gale.
Build me a single HTML page: a person holding an umbrella in wind that keeps getting stronger.
data: URIs are fine.localStorage, sessionStorage, cookies or indexedDB — sandboxed frame, they throw.html block.One paragraph, before the file: what the canopy is made of in your code, what inversion is as a state change, and what decides when it happens.
The single fenced block.
htmlWhere it looks right and where it looks wrong. Everyone who has held an umbrella knows what one does in wind, so say what yours gets wrong before I find it.