RAE research in the 80s re noise abatement / MLS, using two segment approaches (GS) and segmented track offset (AZ) concluded that the final stabilised straight-in criteria using conventional instruments were 300ft/1nm VFR, and 1000ft/3nm IFR. These were the amalgamated results from airline operational assessment, generally in simulation, and RAE research flying at several locations and a range weather conditions.
The largest track angle was 90deg provided the final turn was completed before the straight in point; Max GS 6 deg becoming 3 deg.
One aspect affecting the outcome was the ability to identify the runway at decision height and correct if necessary. The ‘Calvert’ C/L plus 5 cross bar approach lighting pattern provided better cues than a C/L and single crossbar, both for awareness and judgement of lateral / vertical deviation, and aircraft attitude / manoeuvre margin relative to the runway.
Simulation of enhanced visual devices for a head down display and landing indicated that the sensor resolution at that time (1990s technology IR, radar) was insufficient to judge a visual-like flight path, thus FD guidance was required.
There was little difference between an EFIS head down landing using a FD and a HUD FD without visual enhancement; pilots just flew the FD.
A head down display of an enhanced or synthetic ‘real world’ was very sensitive to EFIS pitch / roll scaling which required a change in depiction / ratio of the conventional pitch attitude scale; thus HUD has the advantage, providing the electronic overlay is sufficiently accurate with the real world (value not identified).
Modern technology probably offers ‘Simulator’ type quality, but would simulators have the appropriate level of integrity for flight use, and would each pilot have an independent (dual) display.
Then in training would the pilots be flying a ‘simulation’ of the real world in a simulator, and when using HUD which simulation would overlay the other simulation - a mental conundrum.
The cryptic point above is that when using ‘synthetic’ computer-generated scenes to visualise inside an object people have attempted to place their hand through the structure to reach what they perceive as ‘real’ opposed to the real world.