If StackOverflow jumped off a bridge, would you jump off one too?
You can't base a huge decision like that on "well it worked for somebody else". How did it work for them? What was their platform? What was their data profile? What were their requirements? What was their acceptable level of risk? It could be that they have so many caching layers and lose so little money that they are fine with the database being down for 5 days while they recover. Or that it doesn't actually work that fine for them, because who wants to broadcast that their system is shitty? Or that they've just been lucky.
It's not worth finding out the hard way. Build the best system you can with the time and money and expertise you have. Don't cheap out just because you think you can get away with it - especially on the critical stuff.
You can't base a huge decision like that on "well it worked for somebody else". How did it work for them? What was their platform? What was their data profile? What were their requirements? What was their acceptable level of risk? It could be that they have so many caching layers and lose so little money that they are fine with the database being down for 5 days while they recover. Or that it doesn't actually work that fine for them, because who wants to broadcast that their system is shitty? Or that they've just been lucky.
It's not worth finding out the hard way. Build the best system you can with the time and money and expertise you have. Don't cheap out just because you think you can get away with it - especially on the critical stuff.