Kerling, can you give us references to support what you're saying there. My understanding is pretty much the opposite - that research is NOT showing large safety benefits from large amounts of cockpit automation, whilst CRM training is providing substantial safety benefits.
For example, most analyses of the Kansas City MD11 crash indicate that good trained CRM was a major factor in anybody surviving that, whilst poor CRM is universally shown as a major factor in the Kegworth 737 crash happening at-all. Look up G-DUKK in the current AAIB bulletin, which I'd personally label as a single pilot operational fatality, that was entirely down to CRM failings (call it airmanship or self awareness if you prefer).
On the other hand, there are an increasing number of official publications indicating that increasing cockpit automation, multi-function displays and so-on are not providing any safety benefit, and indeed may be counter-safety, particularly because pilots haven't had sufficient training in how to integrate those tools into their flying. This clearly is talking about GA, because in the airlines pilots are trained in this task integration - they call it CRM !
G