Hi Grant,
There are several good reasons why beam splitter guiding was abandoned in favour of the more expensive mirror slit guiding option by all the other manufacturers (SBig, Baader and Shelyak). Focusing and positioning the star on the slit is absolutely critical. You cannot guarantee long term the splitter image will remain parfocal and in the perfect sub pixel alignment with the slit you need and realignment is problematic working effectively blind, particularly with a spectrograph like this showing astigmatism where you have no idea if the star is in focus on the slit for best throughput. (Terry stated in the video that the XY slit position they supply with the spectrograph could not be guaranteed and the user would need to tweak this to place the star precisely on the slit but this is far from straightforward in practise.) The symptom is low throughput which was seen by users of the LHIRES copy designed by Ken Harison for example and eventually replaced with a mirror slit guider in the commercial version. (This also failed commercially for other reasons). Additionally when guiding on field stars, these need to be 2-3 magnitudes brighter compared with a mirror slit because of the 10% split
Cheers
Robin