Best approach for storing and updating aggregate data

[pahnev]

Hello GRDB community, new user here! 👋

I'm looking to use GRDB to store data from multiple sources in a single table. This is because I'd like to observe the data and its changes in a single view.
The remote data sources have the same "id", helping out to figure out which row to update.

Now during the updates I'd like to keep the values that have been inserted already to the database. What would be the best approach for this?
Should I fetch the object from the database and update that during the import round? Should I aggregate the data in advance and loop over it only once with an upsert method? Or should I simply store them in separate tables and observe each in that same view?