Updates to Shapelet.fitImage functionality

[rmjarvis]

The ShapeletFitImage function currently still takes a pixel scale parameter for the WCS. It should be updated to take an arbitrary Jacobian matrix to describe the WCS. It's not a very hard change, but it was big enough that I didn't want to do it as part of #364, which had already grown enormous in its scope.

While I'm at it, I think it would be nice to flesh out the Shapelet stuff in GalSim a bit more to bring over some of the other shapelet code I have. Things like fitting the deconvolved shape given a shapelet vector for the effective PSF, and maybe even fitting one solution using multiple images.

It's all code I already have, so it would be nice to incorporate it into GalSim.

[rmandelb]

That would be great! The original set of routines in PSFCorr.cpp included deconvolution via shapelets, but we had to remove it because it relied on some NR code - and it did not have simultaneous fitting, anyway.