Remember that time in feb 2005 when Wikipedia went down for several days and lost about a week of data because there was a power failure and they were too shoe strings to have proper back ups, and also had disabled a bunch of ACID features in mysql to get the last bit of performance out of mysql causing the power failure to corrupt the db, because they didn't have enough money for more servers and were desperate to try anything to get the last bit of performance out of their servers?

No? That's because it doesn't happen anymore now that WMF has a reasonable budget.