No, I think there's about a 75% chance a test will succeed, but even a 25% chance of success would be worth testing. There is essentially no risk to a test, and essentially no cost if IB already has the capability.

On your other point, a negative outcome would be a large positive correlation of the "vote" totals with the snarkiness of posts. That to me would qualify as a disaster. Now that I've said that, I have provided any opponents of low integrity a means to sabotage the trial, but such people don't need me to point out what will be obvious to them.

Another negative outcome would be if members find the feature useless or nearly so and it consumes significant screen space. (I don't yet know whether IB's implementation uses any screen space that is not currently empty.)

Another negative outcome would be if posters complain that their good posts are getting no feedback while others' snarky or otherwise deleterious posts are getting positive feedback.

I think it's unlikely to be a difficult decision to call the test a success or a failure. Failure will be especially apparent. In the absence of failure, a member survey will tell us the answer.