I'm about to test with another 500W no name PSU. It has what it calls a "Compatability Switch" on the back and can run at two different voltages. One setting sets the rails at 4.97V and 12.32V measured with a dmm. The other voltage set is 5.13V and 12.7V respectively. This may prove interesting.
Another problem I have realized *cries*:
While testing both the Rhycon and this no name psu, at least once, two of the HDs didn't show up in the controller POST screen. One is the secondary master and the other is secondary slave. Now, this whole time I've been suspecting there may be an intermittent loose power connection (though I did have my doubts). Now I think I've finally realized what's going on. The Highpoint IDE PCI controller must also be shot. Here's why. I have "Y" power splitters going to my top two drives and my bottom two drives. Top drive is Pri. Mas., 2nd drive is Sec. Mas., 3rd drive is Pri. Sla, 4th drive is Sec. Sla. According to the post, its not detecting either of the secondary drives when this happens, yet each secondary drive is on a seperate "Y" splitter". So since the thing in common is that they are both secondary drives, NOT, that they are on the same "Y" splitter, that must mean that the IDE controller is to blame. Right? It can't be coincidence to seperate molex connections come loose at the same time, can it? Yet, when I reconnect power cables, the drives are (usually) detected properly. Maybe I'm jiggling the IDE cable? Doubt it, but I'll still try a new one. Could this all still somehow be just a bad mobo? (Wishful thinking) Am I missing anything? Is my logic flawed?

See why I want to scrap this? I'd save SO much time in the long run. I figure it would cost me around $435.
2x250GB HDs $240
NF7-S $86
Antec 480W $8 (free under warranty + shipping....hopefully covered by warranty)
RAM ~$100
At what point do I give up? At what point do I fork over this money? I should be able to cover the amount, but it's going to be close.