@Jasper_Janssen The 5A positive lead, going from the polyfuse to the upper right hand corner on http://reprap.org/wiki/File:RRD-RUMBA_PCBLAYER.PNG to power the RAMPS, it also branches off to the left to a…Diode? (SMC block that says M7 on it.) Here’s a better picture: http://www.flickr.com/photos/33743995@N00/11778749106/
Paging Dr @Thomas_Sanladerer
(The damaged trace is the one supplying only the rearmost stepper driver, i’ll assume it was connected to a motor and enabled)
First of all, it is entirely possible to damage a circuit by applying less voltage to it - in our case, the stepper drivers are step-down converters, which, under normal conditions, convert a high-voltage, low-current supply into a low-voltage, high-current output. As the supply voltage drops, the drivers starts drawing more current to maintain the same output current (and therefore, voltage and power). Going from 12V to 5V makes for a 2.4x increase in current draw, which results in a 5.8x increase in trace heating.
However, with any available copper thickness (and assuming it was properly manufactured), the trace should have barely stood up to the ~3A that the driver could have drawn. Now, if that nice polyfuse was on the verge of tripping, the available voltage will have dropped even further, again increasing the driver’s current draw. At this time, the fuse should have gone open-circuit (and it maybe has), but it was either too slow or, after that, returned to regular operation in some unfortunate way.
So, to sum that up: Your board might be entirely healthy (except for that trace, of course) and what you’ve seen could have simply been an effect of the reduced supply voltage. Connect a jumper wire from the polyfuse’s pad to the solder point (bottom side) of the driver’s header closest to the corner and see if that helps. And get rid of the polyfuses.
What is somewhat weird, though, is that the trace shows stronger burn marks in the regions closer to the input and looks totally intact as it gets close to the stepper driver. This could be the result of uneven electroplating at the factor, which could be grounds for a warranty claim.
I’d already run wires to the pad near the top driver (it’s actually one trace that powers all of the drivers in parallel…that particular driver wasn’t energised, but that’s not to say the others weren’t…it happened on initial powerup. )
Alas it also took out the diode(?) just above the polyfuse. It’s labelled M7 and has a polarity assigned to it (white stripe on one end) powering up the board caused a pop and smoke at that component. At this point I gave up. 
@Mike_Miller thanks for clarifying what that trace is actually used for - the non-proprietary documentation of the Ramps is somewhat useless in that regard and i don’t have Altium Designer at hand right now to open their source files.
So with this new information, it is pretty obvious that that trace is horribly undersized, even for operation at 12V. RRD claims compatibility with the DRV8825, but if you actually use the full amount of current the DRV can push, you’ll easily fry the Rumba.
That, along with the issues @Tim_Rastall was having, all points to an incompetent engineer and sloppy design validation.