The addition should be designed around the ocpd at the origin. Take your point about overload, but isn’t disconnection about in part safety to individuals?
I still go back to my points above, why is a fuse downstream of something ok normally, but not in this case of a 13A style fuse.
The ocpd at the origin is really the incoming (80A bs88) fuse, do we need to ensure all circuits out of the cu operate this fuse within the 0.4s?
Or do we ensure that the ocpd for each circuit is the one to operate in 0.4s?
Obviously we prefer that the downstream ocpd operates first to give selectivity, if this fails and the bs88 fuse goes anyway that's great - but this latter bit doesn't have to operate in 0.4s like the circuit ocpd.
A downstream fuse for the circuit ocpd is just the same, I would much prefer the fuse (plug or fcu) to operate ahead of the circuit ocpd, and that to operate ahead of the incoming bs88 fuse.
Providing each ocpd disconnects within a suitable time for the section it protects - to me that is ideal
The 13A fuse ensures 0.4s downstream of it.
The circuit ocpd ensures 0.4s downstream of itself at least to the 13A fuse (further is not always a bonus, because it could hurt selectivity)
The incoming bs88 fuse ensures 5s downstream of itself at least to the circuit ocpds (further is not always a bonus, because it could hurt selectivity)
I just can't see any disadvantages hence why I am struggling with your reasoning, if I understood, it could change my viewpoint.