big_omega implementation disagrees with published formula

[translunar]

I haven't actually been able to get it to produce an incorrect result, but neither can I find my derivation of it. I don't want to fix something that isn't broken.

The version from Valenti 2015 is here (but note that it is missing a one-half in Eq. 39):

From Markley 2003, it's as follows:

Note also that Markley gives the quaternion in qx qy qz qw order, whereas Valenti is qw qx qy qz (like pyquat).

Unit test code is attached showing same results whether Valenti is used or the existing version.

I've opened this issue so other people know it's of some concern. If anyone can show an example of it misbehaving, that would be outstanding.

Traces from: https://github.com/autumnsault/pyquat/issues/9