BGG
I am not claiming to be a safety engineer, and have not been an airworthiness signatory for some years now, but I understand your argument.
However, these numbers are being applied in the real world for new designs as an assumption, in lieu of justifiable evidence. As I mentioned in a previous post, I have never seen the assumptions challenged when historical data is available. I personally prefer the approach S/W implies; if you remove the human element, does the system stack up as safe. If not then a critical appraisal needs to be carried out. In my experience, this rarely happens.
regards
retard