One of the major motivations for this is to prevent the deep ground from chilling. We are defrosting it in winter and charging it with additional heat in summer.
During the winter of 2009-2010 I suspect that my measuring may not have been as rigorous as it is now, because of the more frequent oscillations, but the trend is quite visible. The sharp deeper troughs are probably when I didn't give enough time for the ground to recover after turning the GSHP off.
Since the sunboxes were installed in mid March '10, the ground temperature recovered quickly to a higher point at the end of July '10, and has been pretty stable ever since. With the weather improving in Jan and Feb'11, the temperature has actually recovered slightly. I would dearly like to see the summer peak rise higher than in July '10.
When there is demand from the GSHP for heat from the Sunboxes, there is an average delivery of about 1.80 kW from them, so there is still regular heat coming in, even if the Sun is not shining daily.