That usually happens if you have more than one instance of the config app open. You update with one and subsequently update with the other. IoTaWatt recognizes that the update is to a stale config and rejects it.
Regarding the query of Jan 1 ~14:45, it appears there is no data available in the current log. What you are seeing is data from the history log only which is one minute data. The influx uploader only uses the current log, so it sees no data for that period. Query will use either, and so that’s why you see the data in Graph+ which uses query.
Your message log does not show any problems during the outage period, and it appears that it resumes more or less spontaneously on Jan 6. That leads me to believe this has something to do with the SDcard. Your current log is wrapped, which means it doesn’t grow or shrink anymore, you should see the start time advance as new entries are added.
Your history log has 3 1/2 years of data, so I would recommend replacing the card. You can try just copying the old card to a new card and see how it works. You may encounter an error copying /iotawatt/iotalog.log (the current log). If so, delete it from the new card. IoTaWatt will create a new one starting from now.
You may get lucky and whatever you use to copy the card contents is able to read the bad sectors. Any evidence of problems and you should discard the current log.