My suggestion is that it be used above FL200. I know it doesn't help with crossing traffic, but with crossing traffic, you have to be really unlucky - to have the fatal encounter at that exact moment the two aircraft cross tracks. With opposite direction traffic, you're both there on exactly the same track for hours on end.
Australian ATC have assured me that they would have to completely re-draw the airways to implement it, but elsewhere in the world, ATC aren't too bothered by it. Crossing the Atlantic using NATS tracking, possibly one of the busiest and most critical areas of airspace in the world, allows up to 2NM right of track without informing ATC (but rest assured, they can see you're doing it). It's just that so very few pilots use it.
My last 20 years of flying was mostly outside Australia, and flights in Europe, Asia and North and South America quite frequently operate on the same route in both directions. It was my habit, particularly on climb and descent, to kick in a right offset the moment ATC informed me of opposite direction traffic (that I had to sight/pass before being allowed to continue the climb or descent). This not only gave an added safety factor, but assisted in seeing the opposite direction traffic because of the relative movement, which was minimal to zero if the traffic was coming right at you.
The tracking tolerances of enroute airways are such that 1NM offset if used by any GPS-equipped aircraft would not put the aircraft off airways. I am not advocating the use of offset in terminal areas or if direct tracking.
But as I say, there seems to be enormous resistance to it.
Change of subject: someone can (will) correct me if I'm wrong, but I understand that this particular incident was caused by one aircraft being cleared to climb to 2000' below the opposite direction traffic's cruising altitude. However,the PiC of the aircraft cleared to climb did so in VNAV or FLCH, resulting in such a high rate of closure with the other aircraft that the TCAS quite correctly saw the rate of closure as being inside its RA parameters and gave the correct warning to stop the climb. Offset tracking would, in most (but not all) cases prevent this, but an even better fix (and one that simply had to be used in Heathrow airspace) is to use VS mode with a lower rate of climb inserted to give the TCAS vertical mode a set of figures that would not result in an RA.
The purists worry about the use of VS in case someone forgets they're in a degraded level of automation and leaves it in in a situation where the VS the pilot has inserted is greater than the RoC that VNAV or FLCH is demanding. My reply to that is that, while I understand that flight departments have to write SOPS to cover a worst case situation, if they don't allow a pilot the latitude to use his aircraft to best advantage, they're doing themselves and their pilots no favours. I have some sympathy for the poor buggers writing the SOPS, particularly since today we have lawyers lining up to find fault at every step, but the use of VS in these circumstances has a lot of advantages - advantages that I think override the "maximum level of automation at all times" argument.
I think the use of VS in an all-engine overshoot/missed approach (after the initial climb is established) has similar advantages - it slows down to a manageable rate what for many pilots is a very busy (and all too often one they screw up) procedure. The recent Air France incident at JFK comes immediately to mind.