Timelord,
I don't believe the Ben Macdui case makes any reference to the concept of 'theoretical base of solid radar cover'. It hinged entirely on the minimum terrain safe levels and the division of responsibility between pilot and controller for avoiding the ground. None of the scenarios posited by the OP involve descent below minimum terrain safe levels; the variable he/she wishes to explore appears to be the quality of traffic information required during an IMC descent (less than solid radar cover = less than complete traffic information). We know that midair collision risk has been a key focus for duty holders in recent years so I'd be surprised if there isn't already some official guidance material available.
My 2p worth: if timing was critical and I was expecting an IMC descent I would plan my route through an area of low terrain safe level with good radar coverage if at all possible. Inclusion of radar vectoring charts and radar coverage plots in flight publications is therefore extremely valuable. Whether I would re-route to stay in good radar coverage rather depends on the circumstances. In general, I tend toward the view that remaining on track and descending below solid radar cover is a better option than rerouting, adopting 'big sky theory' (which after all is the regulatory basis for the UK's Class G airspace: if the risk or consequence of collision is higher in a particular location then for God's sake do what the rest of the world does and define some airspace other than Class G, rather than introducing non-enforceable local procedures!). TCAS mitigates the risk substantially, but I would be more circumspect if over a glider site without FLARM. I would be even more circumspect if flying air cadets on an AEF. On reflection I think 'best practice' is for aircraft operators to determine based on their particular requirements and risks. There are probably too many variables in this for controllers to do anything other than respond to pilot requests.