Something to keep in mind is that dihedral (or anhedral) only has an effect provided the aircraft is in a sideslip. That sideslip can come from the airplane banking and not moving into a turn courtesy of added elevator or it can come from a yaw courtesy of some rudder application.

Campgems, you provided a diagram above that is classic but sadly flawed. Technically the diagram is right but it would take a very special coordination of controls to bank the airplane and avoid a side slip. It's the side slip that produces a LOT more rolling effect than just the simple geometric change that the diagram shows. An airplane in a side slip that has dihedral will see a change of angles of attack in the two wings. The "leading" wing (think yawing here) will see an increase in the angle of attack while the "trailing" wing will see a decrease. The resulting change in the lift on the two wings will produce a strong rolling force. The higher the dihedral angle the stronger the rolling action due to the larger change in the angles of attack from side to side and the resulting changes in lift coefficients.

So that's how many of our modern sailplanes work that have polyhedral. And, of course, any of the models controlled by just rudder and elevator. Or if you're a serious glutton for punishment rudder only models.

But what about our Galaxy or Globemaster III or other high mounted wing aircraft with anhedral? As mentioned they all have at least some sweepback. Now sweepback is more or less known that 10 degrees is worth about 2 degrees of dihedral. So from the pictures I see the sweep on a Globemaster III just about balances out the dihedral. It's about the same situation with the C5. But these are cargo planes where stability is something to be desired, not avoided. The more I think about it the more I think it's about getting the carry through spars out of the way but to keep the vertical locus of the lift more in line with the true center of gravity of the aircraft and the load it carries. Otherwise the pitch response would be quite different from loaded to unloaded. But as I said before this is more a guess than anything else. It may also have something to do with avoiding the usual lift reducing spanwise airflow. Although the Globemaster still has winglets which are intended to limit the vortex formation related to the spanwise flow..... *shrugs*