Since the manufacturer got these logs, I'd expect them to turn every stone and then some, in order to nail this problem and make sure there is no dangerous bug or some that they may have overlooked during testing. Their answer that says everything is alright is strange from my point of view. They could at least try to reproduce the problem, simulate something similar, try under extremer conditions etc. and only then at least say that hard as they have tried they did not find any problem. At this stage, I would have asked the diver to send his unit for further testing, providing him a replacement for the time being.Yes it is, why I mentioned it.
You could also pick an NDL near the lowest of each dive and compare the 2 computers at that same time point.
Between the CNS O2 and NDL data, you would have quite a bit to help ferret out the problem, share with Ratio.
I sincerely hope that after rigorous investigation they find everything is okay or maybe only a faulty unit and that's all.