extractions in the barycentric frame of each fiber

Hi,

In the process of making progress on this issue:
https://github.com/desihub/desispec/issues/2182
I would like to allow specter do extractions in the Barycentric frame of each fiber as opposed to barycentric frame of the center of each field.
In my mind that would not be a big change. (i.e. in the simplest form it'll be just an evaluation of get_spots() on different wavelength per fiber). But before trying to draft that, I was wondering if there were strong objections to that.

PS A separate discussion for next steps in fixing the parent issue is about changes to the sky-subtraction (they should be pretty trivial and can be done with some resolution matrix manipulation, to reflect a different frame of each fiber)