I have not studied the publication that contained the proposal to make these changes so I offer my view of what it intended to achieve.
It does seem rather a pity that the CAA codified their intent and did not just spell it out.
A recent UK research report that examined all UK VFR accidents - by day or by night -which had a Loss of Control (LOC) element, clearly indicated that most resulted from the pilot being in a position where visual cues were not sufficient to permit visual flight; this could have been the result of: a reduction of light sources (or no visual horizon) at night; or entry into low visibility condition during the day (most likely when IIMC).
Although helicopters generally have a better Field of View (FOV) than fixed wing they are inherently unstable. To fly visually, the FOV must contain sufficient visual cues to permit the pilot to control the helicopter.
At night when there is no visual horizon, as the light sources are reduced, control cannot be maintained. Both by day and night, if there is no horizon and the visibility reduces such that the visual cues are taken out of the FOV - for example if the slant horizon containing the cues drops 'out of sight' below the instrument panel - unless the pilot reduces the height of the helicopter, control cannot be maintained (this can be aggravated by a reduction of speed, with the associated nose up attitude, which itself 'reduces' the FOV). (In fact the reduction of height in poor visibility more often results from this necessity to keep visual cues in the FOV than from a reduction in cloud base.)
There are basically two ways to address this problem: (1) the handling qualities of the helicopter can be improved - thus allowing more concentration to be applied to to obtaining visual cues; or (2) flight should only be conducted in conditions where suitable cues for visual flight can be maintained.
Solution (1) is already applied when certification for flight on instruments is sought (clearly indicating that flying without visual cues itself requires an improvement in handling qualities) - this offloads the pilot's workload and, even in visual flight, would permit reduction in the visual cue environment to be tolerated.
Solution (2) can only be applied subjectively because it depends upon: the equipment contained in the helicopter; the skill of the pilot (the more skillful the pilot the less visual cues are required for visual flight); the FOV of the helicopter; and the amount of workload that is required to perform any task (navigating for example).
It is solution (2) that is being sought by the change in regulatory language. "With the surface in sight" is meant to imply more than just seeing the ground immediately beneath the helicopter out of the side window; it is also intended to imply a consideration of all the elements in the previous paragraph.
In addition to the considerations above, there is nearly always an underpinning visibility requirement - sometime tied to license qualifications, and sometimes to task.
Jim