Checkboard hits the nail on the head.
This is a pretty simple system that has been thoroughly convoluted by terminology. The problem lies with the relationship between the terms Category IIIA, Category IIIB, fail-passive, fail-operational, LAND 2 and LAND 3...and I'm sure the Airbus and other equivalents. Apparently no one at Boeing foresaw the human factors train wreck that was set up by placing the term LAND 2 in proximity with CAT 2. Nor did anyone at the FAA or ICAO foresee the absolutely total intuitive disconnect of a term such as "fail-passive"...although the intuitive capacities of those august organizations has been measurably elusive. Oh, well...we're stuck with these terms.
I will submit my opinion that the terms CAT IIIA and IIIB are useless in anything but a historical context. They did, at one time, represent a staged evolution of the all weather landing concept. There was, at one time, a correspondence between CAT IIIA and a fail-passive system, and CAT IIIB and a fail-operational system.
That is not true today. Fail-passive systems can be used, within certain constraints, on an approach that falls under the Category IIIB definition. All runways in the States are certificated as either CAT III, CAT II, CAT I or VFR. In the EU-OPS, Appendix 1 to OPS 1.430, paragraph e(3), contains a note which says:
In the case of a CAT III runway it may be assumed that operations with no decision height can be supported unless specifically
restricted as published in the AIP or NOTAM.
The note makes no reference to IIIA or IIIB.
The EU-OPS go on at great length, when discussing visual references, to distinguish between IIIA and IIIB operations. I would submit that you could remove these terms completely, and refer instead solely to
fail-passive operations, which by definition require a DH,
fail-operational operations with an authority-mandated DH, and
fail-operational operations with no DH.
OPS 1.430, Appendix 1, suggests a set of requirements that follow logically and which may shed more light on this issue:
(2) Decision Height. For operations in which a decision height is used, an operator must ensure that the decision
height is not lower than:
(i) The minimum decision height specified in the AFM, if stated;
(ii) The minimum height to which the precision approach aid can be used without the required visual reference; or
(iii) The decision height to which the flight crew is authorised to operate.
(3) No Decision Height Operations. Operations with no decision height may only be conducted if:
(i) The operation with no decision height is authorised in the AFM;
(ii) The approach aid and the aerodrome facilities can support operations with no decision height; and
(iii) The operator has an approval for CAT III operations with no decision height.
Notice the references to the AFM, operator approval and flight crew authorization. There are no references, in these paragraphs, to IIIA or IIIB.
From this, and the similar logic presented in the FAA Advisory Circular 120-29A and 120-28D, which covers the topic in US operations, there is a set of questions that I believe may provide a logical way of thinking these issues through during day-to-day operations:
First, what is the airplane certificated and equipped to do? This one is strongly influenced by the MEL and inflight discrepancies.
Second, what is the runway and approach system approved and currently equipped to do? This one is strongly influenced by NOTAMs.
Third, what is the company approved for?
Fourth, what are the flight crew qualifications?
The latter two questions are most likely bundled together in your ops manual, and you may not be able to distinguish which is which.
The bottom line is that company approval can be trumped by equipment deferrals and failures in either the aircraft or in the approach system. Conversely, the company approval itself may preclude utilizing the full capabilities of either the airplane or the runway. Discussion of the above four questions will make for interesting debate during training. I suspect that in many cases, you may find training staff who are also not very clear on the questions we have been debating in this thread. That is a much larger issue.