Bidule, we should not blame certification authorities for a judgement of risk which has been accepted world wide and shown to be be adequate for many systems - and in other industries.
A rule change at this level would probably generate more automation vice attempting to quantifying the variable human in a risk assessment.
The critical point in this safety event is that it happened at all - obviously. There was a component malfunction - three simultaneously, which by all available means had been shown to be adequate. Meeting the requirements at the time, requirements which via higher accountability were acceptable to government and thence the public.
If change is warranted, then consider the process; which apparently has been improved in the icing requirements in CS 25; greater emphasis on ice crystals.
Also, in addition to the mandated modified probe, the manufacture has improved ADC logic, switching, and crew interface, yet some views cite this as an admission of prior error. Operators have reviewed checklists and training, but these and modification are obvious fallacies if used to justify uneducated criticism, which could generate a reluctance to improve safety - ‘we meet the regulation’ - ok, no more.
The industry requires a learning culture to progress safety, a willingness to advance beyond the minimum standard; the question is if increasing litigious action will hinder developing a safety culture - not change rules.
The Certification Authorities should appear for the defence.