Don't get me wrong, I think it's very cool that they're doing this kind of work, and I hope it takes off. What I was really responding to was the idea that this approach would "crush" a more traditionally curated database. At least in the short term, I don't see that happening, because the number of really high-quality data sources is still pretty tractable, because a lot of what's on the web is just going to be echoing/distorting those primary sources anyways, and because trust is such a huge part of that market.
The issue is though that Wolfram is not going to provide the mudane sort of data that people want in the instant way Google are going to claim to do it.
Search camera on Wolfram and I expect you get lost of data on the history of camera's and other such stuff. Google seem to be offering to provide a list of camera models with some pertinent data for each... I suspect the latter will be more "useful" (especially as Wikipedia would probably be fairly reliable for the camera background...)
I think your point is valid: a curated database will be the #1 source for info on camera history and facts. Unfortunately that means Wolfram is competing with Wikipedia not Google... and that is probably even worse for them :(
You make a really good point. For more consumer-oriented data, a more casual level of accuracy is probably acceptable, at least acceptable enough to hang ads off of. I had my head stuck in the world of data that people are actually willing to pay for directly.
