Fixed fl cutout

Previously, the fluorescence cube started at z=0 for all optics options except rms_f50d13 for which it was changed to z=-8. This goes wrong for the infinity-corrected module. I've now switched it so that the cube is mounted a fixed distance (11.5mm) above the camera sensor.

This required a modification to the outer shell of the optics module - but at least for now it looks good to me. @samuelmcdermott could you take a quick look in case I've missed something obvious? It's worth pointing out this will almost certainly be superseded by what you're developing, but I thought it was worth doing as it ought to be a fairly simple fix for some issues that were reported on the forum. You may want to copy over the logic of how the beamsplitter is placed, to get a consistent height.