Centaurus, your thread title overplays the training aspects, as does the discussion relative to the background EASA proposal:
NPA 2017-06.
The NPA is directed at CA 25 aircraft and manufacturers, there is little training content if any, excepting using new/updated systems, monitoring more automation, and managing what could be greater complexity in operation.
The NPA is based on past accidents where the use of lower GA thrust and modified autotrim operation may have provided alleviation. Although EASA has already discussed this with manufacturers, the benefits appear low especially as there is no clear cut cause; other than to blame the human via more training required with existing systems (manufacturers view?)
The referenced
ASAGA study actually makes a better case for training, in particular which instruments to use, and when, and knowledge of critical selections and annunciations.
Pitch, Power, and
Speed opposed to just the first two; not more training, just better training of the key aspects.
EASA GA training is in
SIB 2014-09 - startle, tunnel vision, communication, and management.
I suspect that the manufacturers will have fun with the part 25 NPA, and doubt that there will be any real activity. Cost will be a driver, particularly when judged against a ‘few’ accidents which involved the targeted features and the difficulty proving that these dominated human behaviour, as much as we might wish to believe that they did.
EASA gets rather tangled up with somatogravic illusions. The dominant component is longitudinal acceleration which tends towards a nose down pitch input; however the speed change from approach to GA is generally small, thus the acceleration is low or short lived. EASA also gets concerned about the combination with high pitch attitude and rate, which if excessive might actually be mitigated with lowering the nose due to illusion.
Re simulation; generally training simulators are incapable of generating accelerations and thus any illusion. EASA discusses some interesting research by NLR who have simulated the effect, but only with what appears to be a simulator within a centrifuge !