From https://www.percona.com/blog/2014/03/27/a-conversation-with-...
"we had the MySQL engineering talent we needed to work with the Oracle team to get 5.6 ready for production at our scale."
"We all worked hard to adapt 5.6 to our scale and ensure that it would be production-ready. We found some issues after production deployment, but in many cases we could fix the problem and deployed new MySQL binary within one or two days"
"Performance regression of the CPU intensive replication was a main blocker for some of our applications" followed by a description of how they addressed that.
So it's not vanilla MySQL vs vanilla PostgreSQL. They tailored MySQL to their needs and keep honing it. What they do has little resemblance with what the other 99.9999% of companies do, and I'm probably missing a few 9s. Another excerpt from that post highlights the differences:
"For example, typical MySQL DBA at small companies may not encounter master instance failure during employment, because recent mysqld and H/W are stable enough. At Facebook, master failure is a norm and something the system can accommodate."
My take: if they started with and stuck to PostgreSQL they'd have to work on it as they did on MySQL.