In my personal experience, connecting the reader to the software has cleared the odd issue. I'm not suggesting that replaces any requirement to inform Abbott of the experience to date with that sensor.
Regarding the error codes; I have a feeling that the Abbott telephone analysts probably don't have the lists on the desks themselves. My feeling is they go through a diagnostic decision tree on the support lines with us, being guided at each stage by their script/map. That's a standard approach, usually used to ensure a consistent approach, and consistent outcomes.
To be honest, I use that belief to help me prepare for any calls I need to make to them, to ensure I have the data they are likely to require to hand, including timed/dated examples of "unexpected" readings/conflicts/errors, so that I'm not frantically searching for examples whilst I'm on the line.
I have had several sensors replaced by Abbott, and have never had to request that happen or get into any arguments about it.
It isn't perfect, by a stretch, but it's an extremely valuable tool that I'm sure we each deploy slightly differently and have differing objectives and expectations of.