Gryffin: a large scale web security scanning platform from Yahoo

"At the heart of Gryffin is a deduplication engine that compares a new page with already seen pages. If the HTML structure of the new page is similar to those already seen, it is classified as a duplicate and not crawled further."

Does anyone know what definition they use to constitute "similar"? In particular I'm wondering if you have to do any sort of configuration on Single Page apps which could have remarkably similar markup but completely different behaviors/vulnerabilities.

I don't accept "coverage and scale" as the answer to why this was created. What problem is fundamentally being solved by scanning, or fuzzing, your web based applications "at scale"?

Can someone explain for a newbie on how to practically use it in a project? I understand golang basics. The github documentation mentions what it does but not how to use it.