Where there is discord...
The hoary old Sullenbergers are surely right that adherence to the basic tenets of the profession would have prevented this accident. If there is one lesson to be learned that is it.
But given that every lesson learned may save lives, why not learn more than one lesson? The logic boffins are quite correct that the system handled a subsystem failure in an needlessly unhelpful manner.
There has been much discussion in the past of the risks of automation 'mode' errors. These, from a psychologist's point of view (my former field), tend to lead to 'cognitive misdirection' (if you doubt the power of cognitive misdirection go see an expert up-close magician).
Aspects of this incident bear comparison psychologically with such 'mode' situations. The system took two actions which it takes sophisticated cognition to make sense of - it annunciated 'gear' (directing attention to 'gear') and subsequently closed throttles (without directing attention to 'throttles'). Simply annunciating 'RadAlt Mismatch' would have caught the error upstream both systems-wise and human cognition-wise. That remains a salutary fact, whether or not it's ultimately considered to justify changes to current or future aircraft.