In many cities it becomes tricky to find an actual restaurant. The top ten spots might only contain one or two, plus everyone's favorite ice cream parlor.
In their earliest days, the primary user base is made of enthusiasts. They bring opinions and experiences not available through mainstream critics but with an eye for quality.
At some point, the places being reviewed realize how important the rankings are. They start building things specifically to improve their position on these rankings and you reach a trend-driven, inoffensive mid-market nothingness in the app and in the place itself.
The answer to this problem has always been the same - find and follow a set of critics that you like - keep an eye out for popular food trends and apply a negative bias towards restaurants nakedly following it - start understanding the texture of neighborhoods and the chefs within them: if Alex Stupak opens a new place in Lower Manhattan for example, it would be an automatic stop for me.
But when traveling and wanting to grab lunch, looking around rankings in google maps/yelp you know everything 4-5 stars will at least be decent, which is not useless information.
It even beat out the Night Market, which is the best in the state, if not the country.
I think proprietors and/or customers are able to update their listings and select which meals are applicable. And obviously you'd want to select all of them, even if you are a coffee shop that sells one type of sandwich.
Why?
If you book weeks in advance, they may prepare things differently (in a more costly/time consuming way) or have more staff on site, or make sure that the place is cleaner/tidier than it might normally be.
On top of that, if they don't make it clear that they're reviewers then it'll be an even more authentic experience.