Good point about user error, and making it a conservative test is appropriate, but the product is not intended to be used in that manner. Likewise, people forget to close the lid and engage both clicks on the BearVault or to use a coin to secure a Garcia canister. To be consistent, they should they test those products and the Bearicade with the lid semi-locked assuming user error. Maybe they do, but I doubt they would pass. It would be like testing seat beats that aren't fully latched. Nothing is foolproof because there's always a bigger fool around the corner. The question is, how much should the rest of us suffer to accommodate them?
The history on this is pretty clear, it's taking lawsuits to get a fair judgment on the Ursack. They keep delaying and delaying and changing the rules. They don't even have a test protocol defined yet for this upcoming test.
If these product don't work as intended, we need to know and either improve them or ban them. But do they have to work as unintended as well? If so, then treat them all the same way.