Can we remove avoid_zero_vec(stock_ss(stock__wgt))

[lentinj]

This happens in multiple places, when converting total biomass back to individuals, e.g. g3a_predate_catchability_numberfleet and g3l_distribution.

Ideally stock__wgt would never fall to zero in the first place. In theory this should be possible, but may result in using avoid_zero_vec more than we do at present, if we have to wedge it into renewal, spawning, ...