I think the quibble is that Facebook didn't circumvent Google, they used an open API that Google themselves provided. It wasn't an end-around, it was just Google leaving the keys in the door.

The "circumvent" part is that Facebook's content isn't indexable by Google.

For example, a website review on Yelp winds up on Google, but a review from Facebook won't show up. Insofar as they're creating content in a walled garden that Google can't access, they're circumventing Google.

(It might not be the best usage of the word circumvent, but I still think it's clear what the author is going for here)