RetroWeb says they have bad caps.
Either that or the VRM.

I wouldnt have thought that board would be worth the effort of tracing the actual problem.
Maybe in 15 years time it will be.

Since the VRM looks to be okay, I have ordered an plcc32 to dip32 adapter to reflash the eeprom. Maybe the eeprom died and needs to be replaced? Saw this happen on my acer v65xa board but that one also has an screwed up SIO chip...

While trying to reflash the BIOS I found out that it was okay-ish comparing to the DTK one on TRW with slight differences in string and other(rest are all DOS updaters). Although during the desoldering process I broke some traces. Repaired them after I soldered the chip back on. Still, I've flashed the DTK one on the EEPROM and.... no life...
I am 100% sure it's not the VRM. The VID detection is okay and I get Vcore when the CPU socket is under pressure (confirming the usual case for s478). So now it's either:
1. dead NB/SB (why else would they not work after I plugged in my USB keyboard while it posted?)
2. more broken traces near the EEPROM
3. more loose connections under the CPU making wrong/weak data signal transfer
4. mix of all above

If anybody has an BIOS dump for this board please let me know!