I would agree with your first point Scott, but only because the sample being used is a plain tube, and, therefore, the liquid phase would be serum, rather than plasma. If, on the other hand, the sample was in an EDTA anticoagulated tube, and the liquid phase would, therefore, be plasma, you no longer see in vitro haemolysis, as EDTA will chelate the Ca++, Mn++ and Mg++, all of which are required as cofactors at the beginning of the classic complement pathway.
I certainly agree with your second point, unless, as Anna says, there is in vivo haemolysis in the original sample (which is very often seen in cases of CHAD, even if the sample is kept at 37oC).