Actually haversine in navigation is still evolving.
Hanno Ix and Greg Rudzinski reworked an old method of celestial navigation (Doniol) to get a very compact system with few sources of error, it utilizes ONLY haversines to solve the pzx triangle for both computed height
-almost same as great circle distance for stars if you disregard atmospheric refraction, oblatenes of earth and other real world inconveniences-
and azimuth to the projectionbject altitude With only addition, subtraction and one division.
This acievement does not sound like much, but for a lightly seasick sleepdeprived person in the cramped and frequently badly iluminated confines of a small craft it reduces error sources...