tdracer
Thank you for your comments.
would probably be better for net thrust fuel efficiency to simply close off the inlet to the J58, let all the airflow go around, and make it a pure ramjet
Russian turboramjet experiments of the same era using a MIG-21 engine kept the engine windmilling after transition to a ramjet. They identified the bearing temperatures as being the main problem (needing oil cooling above M3.2 even though there was no friction heat from high RPM and axial loads).
Didn't see the sfc trade offs though.
https://www.cso.nato.int/Pubs/rdp.asp?RDP=AGARD-LS-194