1) Give consumers what they want instead of what you think they want.
2) He used tools as intended. Instead of wasting time implementing Erlang trying to achieve the best messaging platform ever, he went with something that was popular and was easy to understand for everybody at the time, which was PHP. Then went on to improve the system as they needed with HHVM, React, React Native, etc.
3) As much as somebody can hate Facebook for being a social network/privacy leech, which is something a lot of programmers dismiss precisely because of the social part, just the fact they popularized React as one of the defacto front-end frameworks speaks volumes about how they can push real innovation for everybody else.
4) And even simple stuff like sharing pictures. I still remember vividly how awkward was to share pictures before Facebook. Sure, I think Flickr or similar stuff was there already, but nobody is going to keep clicking on your gallery so they can see how an amazing photographer you are. Sometimes the greatest innovation is in the simplest things, and if you ask me, this was THE feature they brought to the table.
It's obviously a conscious business decision.
When a site is focused on "social", like MySpace or Facebook, it comes across as invasive. When a site is focused on personal "marketing", like LinkedIn or dating networks, it can more be useful feedback on your reach.
EDIT: I hate the word "creepy"
This is the first time in my life that I have heard about mixi.jp. Facebook was able to capture not only the US but almost the whole world for that matter. Again, dismissing those achievements as if they were nothing really does not give them proper credit.
Yes, but what he built was yet another banal social media site in php. Nothing that required deep thinking or innovation.