but this all assumes you find no bugs and your design is perfect.
The worst problems arise when you find there IS a bug but can't locate it.
Problems that mysteriously go away tend to mysteriously come back unless you can prove that the change really fixed the problem
100% true. Found myself in this situation several times. Usually at the second appearance of the bug I was doing everything "not to fix it accidentially", but usually I would fail. And once the problem mysteriously goes away, finding it (and fixing it the RIGHT way) becomes near-impossible. Any suggestions against it?

- Check power and gnd connectivity on all Ics and discretes
Connected and short or nearly short. Ground plane encompasses almost all the board. The web of Vcc is present practically everywhere. How do I find WHERE is the short-circuit? (especially if it happens under some chip, or inside a damaged one)? Infrared camera would come in handy...
Also: FIRST point 4 (check if lines are shorted), BEFORE anything else, especially powering it up. Otherwise the symptoms, diagnosis and fix will be way too obvious. Symptoms: smoke. Diagnosis: Fried. Bugfix: Replace everything.

Think about possible �reasons� that can cause the problem seen(Round up the usual suspects)
The usual problem is not that I can't find reasons, but that I find WAY too many.