First, you'll never get users to do it at huge scale, which is what you need. Those days ended - to the limited extent they existed at all - with sites like StumbleUpon declining and with the arrival of the smartphone with apps dominating most average usage. Unlike food (Yelp), the average person has zero passion for the Web and zero interest in taking the time to provide feedback in such a manner (they have as much passion for the Web broadly as they do any other utility; they use it, yes, they get upset when it doesn't work, yes, but they have zero passion for it). To get people to dedicate time consistently, you need passion. The young Web had that because of who the early users were (early adopters have passion), and it was a shiny new toy even to the average person (at least for a few years).
Second, the interface point (for initial rating and discovery sorting) has to be tied into an already existing epic scale discovery point. That's because you aren't going to be able to get people to first browse to the rating site for discovery purposes, the audience for that would be relatively small (which is why StumbleUpon is dead). There are only a few sites/services that could even be considered on the short list for that: Google, Facebook, Reddit, Twitter (and other services in places like China or Russia). You're inherently competing with the huge platforms for discovery. The next huge platform will arrive with an inflection point occurance.
I'm not suggesting you can't build a small service that would be popular with a group of passionate users. I think that's possible. You won't be able to "rate the Web" though, at the scale that would be desirable. You need millions and millions of quality active users for that.