I think the problem is that the is no 'total SQL file' for a DATA sync job. When the tool starts at one end it does not 'know' what comes after. To implement your request we would need to enable a 'simulation mode', where it does not execute commands, but writes to a file instead.
I don't know how useful it is. It is individual INSERTS, UPDATES and DELETES for each row where an operation needs to be performed. It may be thousands of individual statements. And I do not know either how (human) readable it will be. It is the WHERE clause that is the problem. Probably Ritesh can be more specific.
Do you think that you can use that information? I'd rather say it is better to backup data before sync'ing. You can run a backup and a sync as a single batch job.
But let us hear Ritesh …