Coregistration of Sentinel-1 GRD with slice-assembled scenes - blurry outputs

Ah OK. So if I understand the workflow should look something like this:

Currently testing…

Guess this post/thread could be somewhat related :