[Cathar]

Other points of note, a single 120mm radiator with a fan of tolerable noise level has a C/W of more like 0.05, rather than 0.02. Significantly alters the complexion of the pump heat maths. Lots and lots of variables on this one, but a fairly dumb rule of thumb of the pump not having more than about 20W of rated nominal power draw per 12x12cm area of radiator space actually works pretty well.

The only basis of the aluminium vs copper dissipation myth is that based solely upon weight. It is possible to construct an aluminium heatsink that weighs less than a copper heatsink, and dissipate heat better. If we ignore weight and just keep all dimensions equal, copper always wins. The real clincher here though to bring the point home is that the commonly accepted formulas for either passive or forced convectional heat transfer is not dependent upon the properties of the surface of convention, other than the area of that surface. Since aluminium, all sizes being equal, is more costly to conduct heat from somewhere else to that surface than copper, then this is why aluminium is typically worse.