I really don't think comparisons with every other country on the planet will get us anywhere. Countries that do not operate the ICAO airspace structure are all being encouraged to do so, that is why CASA wants Australia to change. We are operating a non-standard airspace configuration and there is no reason why, given our expertise and relatively advanced ATC and Safety Management systems.
Back to the problem:
Currently IFR aircraft entering Australian Class D airport airspace are kept separated from all other aircraft because in many ways that is easier to do than giving traffic on VFR's. It enables the controller to keep better control of the arrival sequence. When they are closer to the airport they are then sequenced visually into the circuit area with the controller providing separation until the pilot is able to sight the aircraft to follow. (This is the simple explanation it is often harder to do because of weather, etc)
In the US the same IFR traffic is being controlled in C and E airspace by the Centre (the only D is around the Towered airport, about 5nm I think) but the process is the same, eventually the D tower takes over for circuit sequencing. If the traffic is RPT and cannot or will not cancel IFR then generally they would be on STARs in bound and SIDs outbound. These are designed to facilitate D Tower separation and the tower has a radar picture repeater to assist with the separation. (Note: Where alternative airports exist such as in the LA area, an airport such as Van Nuys, probably the busiest in the world, will ban RPT traffic because they simply would not be able to handle it)
Unfortunately it is not easy to translate our existing D airport procedures to the current GAAP airports. As I see it the major problem is the lack of an alternative to D steps surrounding them. There is very little airspace room around the GAAP airports so how will IFR coming out of G with the VFRs be handled by a Tower controller?
I think the simplest method would be to introduce Class E out to 20/30 nm from each GA airport, where airspace of a higher category does not already exist. The local Centre or TCU would then have to clear IFR aircraft into that airspace and maintain separation with other inbound IFR aircraft until the D tower could continue the separation. Radar displays for the D tower controllers will be essential for situational awareness.
Quite obviously there is a deal of difference between a regional Class D airport handling mixed RPT, charter and private IFR as well as commercial and private VFR, and a Class D airport where the main users are flying schools. However this difference also exits in the USA and it is the latter where the Tower controllers have the greatest problems in integrating IFR traffic into the circuit area. This is where the pilot who is familiar with the airport and feels he/she has sufficiently good visibility will cancel the IFR flight plan and join the inbound VFR traffic.
Disclaimer: It is a while since I was able to see these things in action (ATC familiarisation at Oakland Centre back in the 90's) so if things have changed I have missed the changes.
Final word, I see no difference between the formal ATC (airways) clearance given to pilots for entry into controlled airspace and the next clearance they get, "Taxi to the holding point RWY XX". They are both clearances. If I was to call a D Tower inbound and be told "Enter on downwind RWY XX report ready for base turn" or" Overfly not below 2000 feet report overhead" then I have an ATC (but not an airways) clearance.