Saving asymmetric eq as a `wout` has some errors

[dpanici]

@SebereX found a probable bug, when saving a DESC eq as a wout and running booz xform for an AR scan of a NAE solution, which should converge to the NAE values as AR is increased, DESC does not converge (which might separately be a problem with high aspect ratio solves) and seems to have a flipped sign for some of the saved B coefficients (the sin theta coefficients specifically)

DESC looks like

While the VMEC ones look like

We dont have a test for wout asym saving so we should add this and check results (load a asym vmec in DESC, save as wout, check quantities agains the original wout within some tolerance)