Actually those values like sel*ret*wt are selectivity times retention times body weight, so are not the weight of the retained fish. I realize that we do not exactly report those quantities you want.
One way to get them is to create a dummy (e.g. nil weighting) observation for body weight-at-age for retained catch of the fleet. SS will then create an expected value for that observation. The only wrinkle is that this expected value will be body weight-at-age', not at true age, because it takes ageing error into account so slightly blurs between ages.
We'll create a SS programming issue to look into reporting this quantity.
Rick