addendum
Hi cfreeman,
Apologies - i missed the word "frozen" in yr post 3. It is unclear whether this applies to all the products mentioned or just oysters.
If all products are frozen, the shelf-life will probably be based on non-safety factors assuming the initial micro./chemical data is within specification. So 2nd part of my previous post will apply.
The classic procedure is to prepare a sample batch of the packed product under representative conditions and do appropriate testing across the labelled shelf-life of, say 2 units every X month(s) for the chosen quality parameter(s) where X may vary with the labelled shelf life and rate of change of parameter(s). The latter factor may be gauged in a preliminary test or by increasing the initial testing frequency. I have routinely used this basic methodology for frozen raw/cooked seafood with shelf lives of 12-24 months.
IMEX the sensorial results are particularly sensitive to factors such as size, raw/RTE, vacuum/MAP, intrinsic product texture/OLQ, glazing style, block/IQF, packaging style, absolute level and stability of storage temperature.
The technique of accelerated shelf life testing (ASLT) can reduce the overall time involved but requires more thought/interpretation (see the file in 3rd link of post 4). No direct experience.
As noted in one of the previous links, the micro.quality should minimally be checked against the specification at end of intended shelf-life. (and more frequently if risk-assessed, eg RTE).