It would be lovely if we could avoid the usually slagging festival that anything like this seems to devolve into. Some of what Dick says has merit, although as usual dick has presented it with all the subtlety of a dynamite tipped sledgehammer.
I'm not really sure whether or not what dick is proposing could actually work, however it seems like there could be some value to giving it a fair go at least? If the technology and infrastructure exists, it really should be used to its maximum capability. I believe the procedure Dick and others have hinted at is known as a radar vectors to cloud break procedure. Its not used in civil ATC in Australia, however the military does use it and it was actually quite a nice procedure. Basically the approach controller descends the aircraft to a RADAR/Surveillance lowest safe until the aircraft breaks clear of cloud and can conduct a visual approach. It has a higher MDA than other procedures, however it could likely be more safe and efficient than a DME/GNSS approach, and you then have the benefit of two people watching out for terrain separation.
As for changes to class E airspace, I think we should give it a try somewhere and see how it goes. It will require more controllers, more training and a change in the way we do air traffic control, however if we can get an appreciable safety benefit then I am all for it. Ultimately when I'm controlling if there is something I can do to provide a better service then I am all for it.