It turns out the lead free solder melted with the heat from the processor!
Most of the concoctions (alloy implies a degree of suitability often lacking with these) have a higher melting point than lead/tin. It is difficult to see how this could occur given the processor would "melt" first.
Possibilities are:
1- The RJ45 socket was under stress ie the hole in the case in the wrong place/socket in wrong place and forced matching. What appeared to be melting was in fact fracture which RHOS is good at.
2- Board assembly is often a two stage process with most components automatically loaded and oven (reflow) soldered. Some big components (eg sockets) are through hole, separately loaded and wave/hand soldered. Possibly some production genius got ahead of the curve and used woods metal or conductive glue?
Combination of 1 and 2
Difficult to understand how any of this could happen as there are only about half a dozen laptop manufacturers in the world (think Quanta is the biggest) and they are not given to sloppy manufacture, waaay ahead of avionics people.