What you need to be sure about, Pace, is whether a specific accident (allegedly attributed to the usage of the IMCR) was caused specifically by the IMCR i.e. was the pilot doing something which he would not have done had he had the full IR training.
The circumstances must not involve illegality, because anybody can do that. They also must not involve a pilot error other than one caused by a lack of specific theoretical knowledge which would have been rectified by the full JAA IR ground school.
That's not a reasonable standard of evidence for this debate.
Pilots are taught not to fly into hills, whether they are IMC-rated or instrument rated. But pilots occasionally make mistakes and do fly into hills. The issue is whether a pilot is able better to avoid such mistakes with full IR training compared to IMC-rating training, which is nominally less than 1/3 of the hours requirement, and if so whether the safety dividend is worth the training cost.
One could in principle shorten the PPL course to an hour so and just tell the student what they need to know to survive having demonstrated effect-of-controls. Any failure to survive can then, by your standard, be attributed to pilot-error, not lack of training. But training is not simply about what is covered in the course, it's about what the trainee takes away and puts into practice.
An accident involving an IMC-rated pilot flying into a hill
is relevant to the debate about whether IMC-rating training is sufficient for operational IFR.