It will also work with 10 rows. ;0) If you need to wire together a pipeline of transforms (e.g. merge 2 datasets->remove empty rows->dedupe rows->filter rows)

It will also work with 10 rows. ;0) If you need to wire together a pipeline of transforms (e.g. merge 2 datasets->remove empty rows->dedupe rows->filter rows) then the visual data flow approach is very powerful. But if you are only doing 1 transform (e.g. base64 encoding) then it doesn't have any real advantages.