is it acceptable to model a 4" weldolet on 14" header by defining the sif & tee (weldolet) at the downstream face of the weldolet (in this case about 9" from 14" header centerline)? or does it need to be defined at the branch/header centerline?

You must define the tee at the branch/header centerline, otherwise CAESAR II won't recognize the fitting nor will the software be able to determine in-plane versus out-of-plane.

What you could do is use the SIF scratchpad to determine the SIFs. Then, knowing these values:

1) Define the first element of the header, checking the SIF checkbox and defining the SIFs computed in the scratchpad.

2) Define the second element of the header, checking the SIF checkbox and defining the SIFs computed in the scratchpad.

3) Use a dummy rigid element from the center of the header to the surface, where the branch starts.

4) Define the branch from this point, check the SIF checkbox and specify the SIF here also.

I think the file attached (downloaded from https://www.paulin.com) would help, at least with some graphics.