This comes up not infrequently in laboratories. Suppose one person, a trainee or postdoc, leaves the lab with his or her manuscript not completed*. Sure, this dearly departed individual may have started the project and/or done the bulk of the work on it.
But still, it isn't a manuscript.
And it therefore isn't going to be a paper, ever, until someone else steps up and does the work. Finishes the draft at the very least. Polishes off the figures. Submits the damn thing. Fields the original criticisms. Marshals the response to review. Creates the revision.
If one other remaining/subsequent person in the laboratory does all this, the dearly departed loses the first author slot. Arguments about the scientific importance of the original idea or the key data pale at this point.
Because if it isn't published it didn't happen.
There is an important practical concern for mentors and you will want to think very closely about this. It opens the door for any subsequent trainee to leave unfinished (as in unsubmitted) projects behind and then later insist that they have the right to be first author when someone else finishes it up. The motivational impact on your trainees' behavior is somewhere damn close to disastrous.
*Yes, there will be some wiggle here about "Oh, I submitted a complete draft to the PI and all it needed was a little editing" when it wasn't even close to being submittable.