I'm neither agreeing nor disagreeing with you. But it's important to realise that this issue occurs with all other filesystems too, including ext2/3.
If the machine has suddenly lost power, then clearly something 'catastrophic' has occured. This is clearly outside the influence of the developers. Yes, they can insert measures to minimize losses and ext4 already does this -> The journaling will give you an intact older revision of the data.
What else would you suggest be done? I'm not trying to start a flame war. I'm genuinely curious.
It's one thing to say that the user is expecting X but if a product is used outside of recommended guidelines, then X can not be guaranteed.
Here's a half-arsed analogy: If I crash my motorcycle because I fell asleep, is that the fault of the manufacturer? Realistically, no.
The answer here is, if your data is soooo important that you can't risk anything going wrong, do the following:
1) Turn caching off. ie: Write directly to the disk. This will kill performance. But at least you can be as sure as possible that the data is written. Alternatively you can manually narrow the cache dump window to something closer to ext3's defaults.
2) Buy a UPS. If the interuption of power is killing your data, just buy an Uninteruptable Power Supply. These things are cheap and, if your data is really that important, you have no excuse not to have one.