Well, I’m in for 2 penny-worth on this one:
I was an active controller (well, a SATCO anyway) during the QFE to QNH back to QFE debacle. The iteration that everyone seems to have forgotten about was that there was a period of ‘QNH for initial and intermediate approach’ and ‘QFE for the instrument final and circuit flying’, which – at that time – replicated civvie practise to a greater or lesser extent; I had to listen to Heathrow Director (the position, not the individual on these boards!) for many hrs responding to ‘localiser established’ with “x miles, QFE is XYZ millibars, contact tower, 118.2” - especially during CAT III operations.
From an ATC point of view the use of QNH simplified coordination greatly with the 2 very adjacent USAF units whose patterns integrated very closely with ours, and the scope for applying the QFE/QNH corrections in the wrong sense was obviated at a stroke; much ‘tidier’, simpler and therefore safer. Coordination with adjacent LARS units became a piece of cake and very quickly Airfield QNH (as opposed to RPS) became the usual datum for aircraft in transit below the transition altitude, which also helped with observance of the ‘set an airfield QNH when below or adjacent to a TMA’ rule.
However, even we ATCOs understood the desire of 'fast-jet' crews pointing towards the ground in East Anglian fog on a PAR or ILS to relate to their height above the threshold elevation – I’d want to do the same! Therefore, I always thought that the ‘halfway house’ was the most logical solution. However, the 2-winged master race didn’t! It was so obvious that a ‘meal’ was going to be made out of the new requirement to re-set the altimeter to QFE after establishing on the final approach track. This was so readily anticipated that I briefed my troops specifically to be circumspect on when to effect the change.
Perhaps the following undercurrents also mitigated against the policy:
a. The halfway house was perceived as a compromise – and who likes a compromise? Compromise goes against the Military ethic of ‘all or nothing’.
b. ‘Blood’ had been sensed when QFE was re-introduced for final approach – therefore, “a couple of CONDOR incident reports of mis-setting the QFE (despite the PAR Controller’s read-back check!), should see the status quo restored, boys! One last effort and victory will be ours!”
As an aside, I don’t recognise what P3 Bellows is on about at all; transit or LARS tracks on the QFE just makes for more mental gymnastics (with the resulting potential for error) when coordinating with other units, whereas RPS and the SAS provided a common datum within a locale; you just wouldn’t DO it as it makes life harder! Unless the rules have changed, the only time that QFE was mandated for use with transit traffic was for MATZ crossings, in order to coordinate with circuit and final approach traffic; otherwise it was RPS or 1013.2, below or above TL, respectively.