Bill -- I had a session with Blundell where I had MaxIm doing automated camera switching etc, and it was up to her to detect which star in the acquisition camera is the one she wanted to measure (not always the brightest). She also has(had) a spectrograph which had a fiber that would migrate around in the field and thus needed to be detected. The Perseus has significant flexure issues based on its load and alt-az position to the point where it needs to be integrated with the mount so it provides flexure corrections. It "almost" worked and then she disappeared.
I have to be honest... I'm unlikely to tackle this project after a couple of aborted tries (Blundell as well as two others) all of which chewed up loads of time and ultimately failed for various reasons not under my control. This will need to be provided with a special purpose set of hardware and software. It would take exotic scripts which I would not be able to support, and probably create a continuous stream of "improvement requests" that would address one or two people. It's outside my business model, especially now where I have a mature and reliable system that provides high-quality results for what it was designed to do (acquire images for science and art), and superior support of student/public usage via the web UI.
If Blundell can show reliable automated spectroscopy, it hardly matters what software does it. It will be a major accomplishment. I would just go with it.