Andew, the following may help alleviate your concerns about manual landing.
The research paper (TR81025) concludes the following:-
Manual landings were successfully demonstrated in RVR conditions lower than hitherto been thought possible. This established the feasibility of safe operations (with high success rate) in low visibilities (down to 240m and below) based on manual landings per se or as the reversionary mode of a fail passive autoland system.
There are additional caveats such as the need for 50ft DH and high accuracy autopilot guidance; these are now requirements in CS-AWO.
The research data supporting the conclusion indicates that manual landings were made in visibilities as low as 150m RVR, although the success rate based on the ability to decide before DH and then complete a landing started to reduce below 220m RVR, i.e. no land decision given, or the landing was attempted, but judged excessive workload.
Note that this data was for a manual takeover at DH; for landings simulating an autopilot failure below DH, the corresponding RVRs were lower, with 100% success rate at 200m and the lowest value 120m.
Note that the piloting task in the tests differed from that in commercial operations. During the tests the task was to detect and assess the conditions and cues as being suitable for landing, and then complete the landing – only the first and second attempts in the conditions were assessed to avoid ‘learning’.
In commercial operations the required visual references are thus already defined, and in stable (Cat 3) fog conditions within the RVR limit they should always be seen. Therefore the crew task is much less and crews will have the benefit of training. This provides commercial operations with an additional safety margin from reduced workload and RVR minima higher than the limiting condition.
One aspect which is not specifically included in the regulations is the cockpit view over the glare shield. As you may realise this is a critical parameter in determining the point of first contact, the extent of the visual scene, and maintaining the view during the flare. This is a judgement issue for the test team, which is encompassed in CS-AWO 321 (b) 1
(A) It is demonstrated that manual landings can be made without excessive workload in the visibility conditions; and
(B) The aeroplane has a low approach speed, is easily manoeuvrable and the height of the pilot's eyes above the wheels is small;