Plenty of lazy wanna-bees do things poorly and put organizations at risk and expense of greater downtime; you seriously make your employer wait for fsck of terabytes of ext3 data?
Sounds like you might be one of those lazy wanna-bees.
First, ext3 is a journaled file system so if you are waiting for a fsck on a it then you probably need to run it. Second, if your in a position where you are having to run a fsck on production data then your an idiot. If the system is so hosed that you have to run fsck, then you take that system out of the rotation and bring the back up system on line. Then you can run fsck at your leisure.
ext is very non-robust, can lose entire filesystem if power goes out at wrong time. And fsck time is enormous compared to superior and more robust filesystems. I administer hundreds of servers, ext only used for boot
I can't recall losing data on a ext3/ext4 file system but I do recall losing it on a rieserfs at one point. Now, if you are in a position where you can lose data in a real production environment because of something as simple as a power outage, then you probably shouldn't be in charge of hundreds of servers. Production servers have dual powersupplies plugged in to independent power sources such as separate UPS. The UPS themselves are only there to keep the systems alive till the generator backups can come on line. Point blank, if you lose data because of a power outage you are a moron.