Kerbal Space Program

Currently, the drag of non-wing parts has a contribution from each face proportional to the cosine of the angle between the airflow and the face normal.
Simple aerodynamic theory would have the total force proportional to this cosine, but of that force the component in the direction of drag gets another factor of the cosine; for glancing angles, the cosine is very small, and the bulk of the force on the surface is body lift.

In KSP, the Mk2 fuselage parts show the difference dramatically. At a 3° angle of attack, the bottom face of the Mk2 parts see the airflow with an angle whose cosine is 0.05. KSP applies 0.05 of the drag that bottom face would feel if normal to the airflow to the craft; one would expect only 0.05² = 0.0025 of the 'belly-flop' drag would apply in the direction to slow the craft.
All parts except those with wing-like aerodynamics are affected; the is a bit more context at https://forum.kerbalspaceprogram.com/index.php?/topic/182616-the-drag-of-parts/

The attached plot of drag versus AoA is made by the informational mod 'Wind Tunnel'.

The drag of wings depends on AoA squared, as expected.

Changing this would disrupt gameplay balance, so I categorized this as feedback only. Players have responded to this aspect of KSP by avoiding Mk2 fuselages for space-planes, or more interestingly by giving wings an angle of incidence sufficient to give 1-G lift as they cross the sound barrier with the nose of the space-plane precisely along the airflow.