The implication is that skills are declining. However, where flight safety objectives require matching skills to the task, the statistics suggest that there is a sufficiently matched balance - we are statistically safe, for now.
Some academic views suggest that aviation is approaching a practical limit of safety*, and that there is little more which can be done without disturbing this delicate balance. The residual, rare accident scenarios cannot be imagined nor trained for; safety is as good as it gets - we must maintain current the status quo; beware complacency.
EASA LoC statistics are dominated by high fatality accidents. For the accidents which originated from equipment failure, modifications have removed the particular threat. So why train for these specific situations - after the fact.
The remaining adverse events - relatively few in a safe industry, are more likely statistical noise.
Future risk depends on data classification and subjective assessments (ERCS). The LoC data are wide ranging, EASA Safety Review;
‘
Aircraft upset, includes all occurrences involving an actual or potential loss of control inflight, which includes situations where unintended deviations from the flight path has occurred. This covers only occurrences during the airborne phase of flight and may occur as a result of a deliberate manoeuvre (e.g., stall/spin practice). It includes occurrences involving configuring the aircraft (e.g., flaps, slats, on-board systems, etc.) as well as stalls on fixed wing aircraft.’
N.B. ‘potential’, judgement after the fact. Risk assumes that the future will be the same as the past.
The emerging risks are more likely to be self generated, over-reactions, thinking that we know better - a belief that skills are reducing. But which skills, whose judgement, which situations, when?
We dislike uncertainty, thus in avoiding this, the greater risk is that our constraining, after the fact, pigeon-hole safety categorisation, and rule based reaction will be mistaken; the next accident will be self inflicted, by whom, where, when, … (how many previous ones were - AF447, 737Max - a point of view), limited by foresight.
‘Risk is the amount of uncertainty humans are expected to manage’.
* The paradoxes of an almost totally safe transport systems: Amalberti
https://www.sciencedirect.com/scienc...2575350000045X