I have been considering a 300 to 400 lines/mm diffraction grating (a good one costs almost as much as the S50!) for this purpose, a good quality optical prism would be an interesting alternative if much cheaper. Can you please provide a link for where you got it from?
If the prism/diffraction grating covers the S50 lens only partially, say 50%, you will get a mix of normal and spectroscopic images, maybe that's enough to have the S50 do the tracking. For pure spectroscopy use tracking is not really needed, star trails are irrelevant, especially if perpendicular to the spectrum. All we would need would be the ability to center on the desired star, then start saving 10 second acquisitions without requiring successful tracking and stacking, simply save the raw frames. This would be more than trivial to do but good luck convincing ZWO to support such a request. They will thank you for your idea and consider it in the next 5 years or so.
