I think the short answer is that nobody knows, because there are competing pressures.
The airlines, according to Eurocontrol, want a purely RNAV environment (which to them means INS, and loads of DMEs all over Europe so their INS works accurately; one Eurocontrol proposal was for 300 new DMEs around Europe) and they don't want radar or any navaids other than ILS. This, it is claimed, will reduce the IFR enroute charges. Eurocontrol seems to have bought into this "dismantle most radar and navaids" way of thinking - at least to the extent of talking about it, running workshops (at which the proposal is roundly rubbished, especially the radar stuff), and publishing proposals.
But the national powers to be cannot simply implement this because there is other traffic "up there" also, plus there are ICAO obligations, etc.
Then there is the gap between published approaches and what is actually flown. At any major airport, there is radar, and almost nobody is flying the published (procedural) approaches, and it is these which use the navaids. Generally, the place has an ILS and you get vectored to the localiser. ATC like it because it gives them total control, and pilots like it because they have very little to do. But there is often training traffic which tends to be banging the navaids. Or the radar could be out of service (because it's broken or due to staff shortages). Or the ILS is only on one runway direction, so the navaids have to be there for when the wind blows the other way (even if, like say Bournemouth, ATC give you radar vectors onto the "NDB inbound").
That's why the death of the NDB has been forecast for the past 50 years
