Did later versions of QJM or other books have a better understanding of river crossings and their effects on casualties ?

The shoreline vulnerability factors are the exactly the same between the original QJM (as puslished in NPW) and the current TNDM. The TNDM table has be expanded to add "Sha" (casualties for attackers) and "Shd" (casualties for defenders). So for example, crossing a major unfordable river under small arms fire, the attacker casualties are multiplied by 1.5 while the defenders are multiplied by 0.67. Across a major fordable or minor unfordable river the factors are 1.3 and 0.76 respectively. Across Beach the factors are 2.0 and 0.50 respectively.

They only appply when "under small arms fire" and do not apply for under light or medium artillery, only the vulnerability factor applies.

Note these factors degrade over time...with the note for across beach being that "Applies first day only: second day 1.67; third and subsequent days 1.33 (as long as FEBA is within indicated proximity of shoreline)." The factors for the rivers only apply on the day of the crossing attempt.

There are lots of little adjustments like that from the original version of the QJM (as there were multiple changes made to the model after the publication of NPW) and in the TNDM.