Gonzo, there is some history from BLEU before the 1965 accident – see ref.
The originating Cat 1 limits were 200ft and 800m; much of this considered the then accuracy of the ILS and/or flight system, where the final approach position often required a lateral manoeuvre back to the centreline. Thus the visibility requirement had to be sufficient to both identify the aircraft’s position and to enable a correcting manoeuvre; these tasks were supported by the Calvert lighting pattern.
Over time, the ILS accuracy improved such that the need for lateral manoeuvring was reduced, and the accuracy of approach delivery further aided by FD and Autopilot use. Then with a process of demonstration by aircraft type (usually the manufacturer) or by operator application, lower visibility limits could be approved by individual authorities (use of UK CAA ‘Fog Model’); the 1965 accident noting ‘the lowest limit’.
Formal recognition of these aspects cumulated in JAR-OPS 1 (E), which with due harmonisation process settled on the lowest values used by all European authorities (previous ref was ECAC Doc 18). This harmonisation actually increased the risks in some circumstances (manual, no FD, poor beam quality), but this was either accepted or more likely overlooked due to the high landing success rate for Cat 1.
Subsequent work by BLEU (RAE Flight Systems Bedford) identified some of the underlying parameters governing the minimum visual segment and decision making, and some authorities have further lowered the limits for special circumstances or equipment use (FD / Autos to 80%DH, aircraft type). Much of this is now being accepted as the norm, which further erodes the approach safety margin vs min RVR.
http://aerosociety.com/Assets/Docs/P...g_Charnley.pdf