Dear tech,
You asked about validating as PRP. My experience includes processing raw clam/mussel meat which typically arrived at the factory mixed with large amounts of shell, bits of rope, the occasional stone etc, etc. Sorting was totally by hand.
For a traditional
haccp analysis, my original plan included the sorting step as a
CCP based on the existence of ideas such as the already mentioned US regulatory comments. This generated considerable humour when audited and a comment something like "no other processors of this product ever include this as a
CCP, for us it is always a PRP, please change it" (the basic principle of minimising total number of
CCPs). So my
CCPs decreased by 1.
(eventually, if the percent by count of shell exceeded a certain level, we simply charged the supplier for the excess labour, more effective than rejection

).
You might summarise the above as reverse of metal detectors. The latter is a typical pragmatic
CCP for audit purposes, sorting IMEX simply the direct opposite.
If you are interested in validation, I'm sure it can easily be found as a non-significant hazard for
various products if you do a little searching. (as example seafood, the US publishes an official list of typical
generic CCPs, no mention of such a sorting procedure). Nonetheless, yr specific risk situation may be different, ultimately,it's
yr HACCP system.
Regarding size limits, I hv seen use in specifications of max. 2mm in Europe / Australia for hard shell and recently 0.8 mm in another document (sorry, didn't note the link).
If out customer had required a nil contamination, the business would never hv started. If you want spec.validation, it is surely available on the IT.
Rgds / Charles.C