Many orgs already gather data. See essential projects and interesting projects. But they don't put it in a format that can be filtered for one-purchase at a time for various people's criteria. That's the objective here. At least so far.
And, yes, you're right, the Signal Wiki will be simple. It will not look like this Development Wiki. It probably won't use the same software, for instance. Ideally it would use what Metaweb creates.
On the UI issue:
There's no reason not to distribute data on what's in the store in book form, even customized book form. You could for instance slide a card into a slot, and get a printed book with price premium information customized to you, or generic information like a score. It might even get printed at the retail shelf someday if some nation decides to require it - the first step to Consumerium Country. There's no bias as to how the signal is delivered. That's the main reason the language IS so abstract right now, actually.
If you think the existing default end user vision is wrongheaded, well, write up some worst cases about it failing. Otherwise there's nothing specific to discuss. Write up some best cases with a book or guide at the shelf. The whole reason to solicit these use cases is to figure this kind of stuff out.
But there are several points of Consumerium Service access, and they would have to key on some product identifier, and the barcode is the standard one. We also know that it gets scanned at the checkout counter, so it might just be a matter of adding up a score, like a game. There might be benefits or a reward scheme for buying moral products. But sometimes just to know your score is valuable. There's merit to treating it like a game, where you are EXPECTED to memorize or just shift habits so as to buy more ethically.
But all of these things would still require someone to barcode scan to be sure what the product was.
And it's a fantasy to believe that companies whose behaviour is exposed by the Consumerium Services as undesirable would not try to interfere legally. We would have to be damn sure of our data. This isn't a Wikipedia type wankfest of amateurs. This data is supposed to be reliable enough to bet on. The whole ConsuML proposal is about drawing on data from many sources to dump into Consumerium:intermediate pages, and hopefully not creating new data that would have to be verified by volunteers here.