It would be interesting to know if either the aircraft involved in the near-crash as described by arc-en-ciel (page 64) or the actual accident aircraft had the QFE pin-program installed.
Two things lead me to ask the question.
1. The report and graphics that arc-en-ciel has posted refer to QFE.
2. Phil Squares (Page 61) appears to have been privy to the accident DFDR data and indicates that an intended managed approach was changed to selected at a late stage.
The Airbus QFE pin-program sees the MDA (ie "baro minimum") field in the MCDU permanently changed to MDH irrespective of whether QFE or QNH is set. When flying a QNH-based selected NPA the applicable MDA is inserted in this field but is displayed labelled as "MDH". However, when flying a managed/managed NPA the associated MDH must be inserted into this field, irrespective of whether the approach is to be flown on QNH to an MDA or on QFE to an MDH. If this is not done correctly the "triple-click" and mode reversion will occur at the wrong altitude. This set-up makes it quite awkward to change from a managed/managed to a selected approach, without having an incorrect MDH in view, with associated erroneous or missing auto-callouts.