Nigel,
I think you have misunderstood the point; it is not a question of 'mind-on-the-job', more a mismatch between what has to be done and what can be done.
VFR flying is undertaken by utilising visual cues to keep the helicopter straight and level; when the visual cue environment declines (because of lack of light and light sources at night, or lack of visibility by day) it takes more processing power to assess the remaining visual cues.
As the proportion of processing biases towards assessing the (diminishing) cues, there is less-and-less available for flying (an unstable platform); this will eventually lead to loss of control.
There are other issues also concerned with reduced visibility; the visual cues can only be taken from the available cues within the helicopters Field of View (FOV), as visibility declines and the visual horizon moves closer (and therefore lower), it moves slowly, but inexorably, out of the helicopter's FOV. This is one of the reasons why a pilot descends - not because of cloud base.
There is an extremely good research paper about this issue which is about to be published by the CAA (don't know what the delay is).
...now I am in the middle of this period of high activity (=mess) and the telephone rings...
Jim