Whether or not you can, or indeed how you should, carry out a 3-axes approach is not the issue here.
The question remains - why would anybody CHOOSE, especially in marginal weather, to perform a 3-axes approach in a 4-axes capable aircraft? Or are we now advocating mixed mode flying?
If a similar coupled approach data set was analysed for the 225 we are unlikely to see even 2% of approaches flown 3-axes, let alone 82%.
So, is it the case, as Cyclic contends, that the L2 "IAS hold was quite poor" or is it reasonable to argue that, as DB puts it, "the IAS and the ALT hold on the AS332 is more than competent"?
The factual information gathered by the AAIB, and presented in their report, seems to support Cyclic's statement. Whether or not they were correct in doing so, for some reason, pilots were voting with their feet and electing to fly an approach in the L2 with only 3-axes coupled, the majority of the time.
This data was drawn from the operator's FDM records and that raises some further questions. If the clearly identifiable, mixed mode, 3-axes approaches were not neccessary, as the autopilot "was more than competent", then you might be forgiven for assuming that these events would be flagged to the training department for consideration and correction. On the other hand, if the evidence suggested that the "IAS hold was quite poor" you would perhaps expect to see some communication with the OEM to highlight the issue, and the associated feedback regarding the fault investigation and intended resolution.
However, if nothing was being done to address the perceived, or otherwise, problem with engaging a 4-axes coupled approach on the AS332L2 Super Puma, and pilots were being left to establish their own "work around" - then some might say, this was already "an accident waiting to happen".
Taking the quoted news headline,
"Helicopter would not have crashed if pilot had......"
I'll let you fill in the many blanks, which I'm sure will also be explained in detail to the inquiry.