When I worked at Marco in Seattle, I was the guy who was both delegated to and embraced doing crap boat stability.  There the issue was how many crab pots could you load on deck and take out to the grounds.  The more pots, the better the profit for that year's crabbing (a very short window of time before the season was closed for another year).  Too many pots and the weather turns?  Capsize time.  A sad inch of news column in the Seattle paper.

I was doing this after the stability regulations had just changed to incorporate the effect of a sudden blast of wind from the side.  Before it was all based on the static stability of the vessel; all stuff you could measure at the dock. This new calc used both the measured and calculated numbers and some crude assumptions thrown in because enough data was out there to show the end result worked well.  Boats that passed regularly came home.  Boats that were lost often didn't.  (There are mistakes you can make running crab boats that wil sink even the most seaworthy.  Plus ice - I ran calcs assuming a large amount of ice frozen high on the crab pot netting but nature doesn't always stop at the numbers set by the regulatory bodies.)

Now, further complicating all this is the level of fuel and water in the various tanks.  These can have a huge affect on stability.  The gift to me was the locally written commercial stability software.  The vessel would be modeled in as a 3-d shape.  Tanks located inside with their 3-d shapes.  The weight and center of gravity (CG) of the vessel and all permanently attached stuff is entered.  Then, all the changeable weights, ie each 600-800 pound crabpot and their CGs.  Last, the program is told the percentages full of each tank.  Program then tips the vessel (say 30 degrees), calculates where the tank liquids would settle to, adjust the fore and aft trim of the vessel to meet the changed CG location, recalculate the tanks, repeat the trim and so on until "good enough" and calculate the stability for that degree of tip.  Repeat at the next angle.  (That program was/is amazing.  Floats the boat like a toy boat a kid held in the bathtub, then let go.  Boat oscillates back and forth, gradually settling in, just like real life.)

This has to be done at each tank fill condition so a lot of cases!  I would then harvest the key numbers for each condition and enter them into the Excel sheet I created to calculate "Severe Wind and Roll".  See if it passed.  If no, remove a few pots and repeat.  If yes and by more than a small amount, add some.  ($$ in the owner's pocket.)  Finally make easy to read charts that show allowable pots with fuel usage, both for non-icing conditions and icing.

And yes, this directly affects both the owner's and crew's annual income and the likelihood of them staying out of the news.  (A poorly done report just gets ignored by the skippers when they know a condition is presented as safe when they know it isn't or they are told that something they've been doing for years isn't.  Good reports too get ignored.  Sometimes those boats make the news.  To my knowledge, the skippers with my reports retired.)