Fix convention of signed angles

[willGraham01]

Is your feature request related to a problem? Please describe.
First reported here.

Several of our methods that return signed angles, notably

• compute_forward_vector_angle
• compute_{ego,allo}centric_angle_to_nearest_point

provide a toggle for the sign convention of the returned angle(s). However the effect is usually just a sign reverse, so for the sake of simplifying the user interface and providing concrete concepts and terminology, we should fix the convention we're using for the signed angle and not provide this toggle in our functions.

NOTE:
The v_as_left_opperand argument to compute_signed_angle_2d might also be able to be removed, though this function is asymmetric in u and v, so care should be taken to check if this can be done. The function docstring contains some more elaboration.

[niksirbi]

The v_as_left_opperand argument to compute_signed_angle_2d might also be able to be removed, though this function is asymmetric in u and v, so care should be taken to check if this can be done. The function docstring contains some more elaboration.

Yeah we might decide that it's worth keeping that option in there. As this is a lower-level utility, doesn't matter as much.

For the other 3 functions you mentioned, which are user-facing, I think we can and should get rid of the toggle.