Sure, I understand that. But using your method as well as the CSV method in the example, I would have 2 SQLite tables when the application is running. I would have to open MyCustData table and traverse through it, adding data to B4XTable1 using SetData etc., then when the user modifies B4XTable1 data it updates B4XTable1's internal SQLite table. When the app is finished I have to write all that B4XTable1 data back to MyCustData table.
I just think this is a lot of unnecessary data copying that could be eliminated if B4XTable could access an existing Sqlite table by specifying what columns you want to appear in the B4XTable grid and update that data directly. Maybe there is a reason for creating a duplicate SQLite table? I don't know. It just seems odd to me.