Result

Everything above rests on a rectangular hull with a rectangular waterplane, which no real vessel has. A genuine hull narrows towards the ends, so its waterplane second moment is smaller than the box figure and BM comes out lower — the box approximation is therefore optimistic about stability, which is the wrong direction to be wrong in. KB from the box or Morrish estimate is likewise only an approximation of a real vertical centre of buoyancy, and both assume upright, even-keel floating with no trim, no list and no hull deflection. GM describes initial stability at small angles only: it says nothing about the righting lever curve at larger heel, the range of positive stability, the downflooding angle or the effect of deck-edge immersion, all of which decide whether a vessel actually survives a knockdown. Free surface correction assumes each tank is a plain rectangle with an unrestricted surface; swash bulkheads reduce the effect substantially and irregular tank shapes change it. The commonly quoted minimum GM figure is informational context, not a regulatory determination — actual criteria depend on the vessel type, the flag state and the applicable intact stability code, and are assessed on the full righting lever curve rather than GM alone. Use this to build intuition and to check the shape of an answer, never to make a loading, ballasting or departure decision.