I’ve just had a quick look at Charles de Gaulle LFPG and pages 30-106 (of total pages 164) are these procedures.
I know they could be flown in case of loss of comms but how likely is that with an airliner, in the terminal phase? Loss of comms does happen but usually enroute, due to a mix-up and at 500kt you are soon out of range.
They could also be flown in case of loss of ATC radar but at one NATS presentation it was stated that capacity would drop to 1/5 i.e. a de facto near-shutdown of the whole system.
These procedures dictate airspace design, too.
I wonder if this is largely tradition?
For decades, Eurocontrol has been talking about “4D” i.e. airliners flying by themselves all the way, at exact programmed speeds, etc. IOW, making arrival and departure routes totally precise. Is this any nearer? Some, familiar with the way Eurocontrol runs, have said it is Eurocontrol unix programmer software utopia 
The RNAV STARS and RNAV initial approach are flown most of the times when it’s busy, imitating a downwind, base and final clearly defined by ALT and SPD at certain points so are kind of 4D.
Exactly, at big airports you often fly them.
I don’t know about LFPG but in Italy I get both STARs and SIDs very often with full transitions while in Croatia it’s very rare to fly them fully – they are usually shortened with vectoring or directs.
