When you run the import script again with just the new data it will run much faster since it won’t be importing the data again. So it’ll take only a few hours. And each subsequent run will be faster since there will be less be data to import.
You can then speed up that by modifying the queries to return only data newer than a certain time. Most scripts I have touched have an import_after setting for just this purpose (but also to allow faster development by importing a small sunset) subset of the data).