Mad as a Mad Thing
I am not suggesting that Class E is the right solution for the kind of areas you describe. As you say, these should be Class D (or stricter), and ICAO agrees with you. My point was that we should consider converting some of the current expanse of Class G into Class E.
All of the issues you describe apply in UK Class G airspace when IFR traffic is receiving a deconfliction service, which is daily business at any military airfield with a MATZ, and around a number of civil IFR airports such as Exeter and Inverness. And in Class G airspace,VFR traffic can legitimately be 'clear of cloud, in sight of surface' with visibility sufficiently poor that high-speed IFR traffic has absolutely no hope of seeing it. At least in Class E you can assume that the unknown VFR traffic is complying with a more stringent set of weather parameters.
I agree with you that Class E is not a known traffic environment. It is not meant to be a sanitised area in which the controller rules the roost, which I know is an odd concept to some UK ATCers
. The reason why I think Class E would be a better way of defining some of our current Class G airspace is that it provides an internationally-understood description of separation standards, service provision, pilot and controller responsibilities, etc - which at the moment are defined on a UK-specific basis through application of FIS in Class G.
I would argue that the majority of UK Class G airspace above 3000' or so is already operated along Class E lines, with the overwhelming majority of enroute IFR traffic getting a service of some description, most likely a deconfliction service if IMC, and the majority of VFR traffic going about its business unfettered by ATC. So why not call a spade a spade? As I've said before, it's not about safety, it's about using the international language of ICAO to describe what's going on, rather than the bespoke language of the CAA.