Yes, the vapor pushes the liquid. But if the volume taken up by the liquid in the system is constant (with ambient temp), which should be the case more or less, since the liquid line is supposed to be completely in liquid form, and then there is a short section of the condensing coil which is liquid. Then the volume of the rest of the system where the refrigerant is in gaseous form is also constant. That means that the average density of the gaseous refrigerant is also constant. The only thing that could change with (ambient) temperature is that the density upstream of the compressor could become lower and the density downstream of the compressor then becomes higher. This in turn would mean that the compressor has a changing pressure ratio, which I find hard to believe for a positive displacement pump.
Bottomline: I still think that the decreasing superheat with increasing ambinet condensor temp is caused by the increase in pressure (not density, remember this is a closed system) that in turn increases the saturation temp. So it's the increasing saturation temp that is causing the superheat to decrease, not the lowering of the gas temp.
This is an intersting discussion!
The funny thing is......
I bet you dont even own a screwdriver.