However, your example image isn't exactly representative of what his coordinates would define, which would be two adjacent pie slices instead of a full rectangle

That really depends upon whether the imagery in question is raw sat data or post-processed.

CCDs don't adjust their shape with reference to the 'latitude' of their current position. Ie. The images they take a always rectangular. Hence the reason why images of the poles (of earth or other planets) are often made up of composites of lots of frames that are all rectangular, but overlay each other at weird angles. Like this one of the Martian South Pole.

The poles are (can be) imaged on every orbit, each time at a slightly different angle as the orbit precesses. Whether the positioning of (the corners of) such images are given in lat/long or some other perhaps spherical coordinate system I don't know. But given that the OP mentions 0 - 360 longitude rather than the more usual -180/0/+180 used by mapping (and GPS?), assuming that the image coordinates have been converted into a Mercator-like system seems unlikely.

...%360

You have to be careful using modulus on float data, it integerises:

print +(123.456 + 180 )%360;;
303
[download]