You’re sequence leaves a gap.  Any changes after the save command and before the journalling starts will be lost. Start journalling   then SavChgObj  then run the job  That gives you two routes to recoverybacking out the changes you don’t want  or restoring the file and applying the changes that you want.

You save an object in order to be able to restore the object. Saves work at an ‘object’ level. It not only includes (all) records at the time of the save, but also includes the record descriptions, ownership, authorities and full description of the object.
 
Journaling works on records, essentially saving single records at a time. It can allow both forward and backward recovery from a point in time. An additional important aspect is the ability to audit procedures. A journal record tells which user (ID), job and program made each change and when each change happened.
 
Saving an object is not required before journaling is started. It is, however, a very good idea. It’s irrelevant whether it’s with SAVCHGOBJ, SAVLIB, SAVOBJ or any other save method.
 
Although not required, coordinating a save of the object with the start of journaling can be very useful. Recovery from system failures or staff/human errors can be much easier if both are available. Errors can happen while running commands that change journal behavior.
 
Tom