I'm confident that I could build a social network that scales with a 3 person team. In fact, I've done it before, but it relied on hosting and data providers and frontend frameworks and languages that I definitely could not have developed the entirety of as a 2-person team.
Do those factor into the count, or are we in the "if you do everything yourself, a cup of coffee actually costs $300,000?" At what point does labor specialization come into play? Because there's the guy who does everything "from scratch" and realizes exorbitant costs for small things, and on the other end of the spectrum is the guy who started a profitable coffee company with $1800 of seed capital, down to the practical reality of the median cup of drip coffee, brewed at home, costs approximately $0.30.
Similarly, sure a 5-dev team probably couldn't build and operate the full-scope of FB (unlike the clone someone could build in a weekend that doesn't take into account hundreds of small capabilities) - but maybe a 500-dev group could, rather than needing 5000-devs. Especially with a narrower focus on scope and outsourcing various problems (such as do they need to operate their own video hosting platform vs partnering).
There are advantages to scale, but also costs, and at a certain point scale tends to be horizontal rather than vertical. More products/features/adjacencies that now need to be coordinated, rather than a solid core and leverging the market for the rest.
Like in software: a 5 developer team is more efficient and has less overhead than Facebook, but a 5 dev team can't build a Facebook.