NUC'ing internally does use a mechanical shutter mechanism. Any moving part is susceptible to wear, but I know that FLIR tests their shutter mechanisms for millions of cycles. Mechanical shutters also draw power. The old R-Series scopes had the option to disable the manual shutter. Nice feature, imho. The newer PTS system is a bit different. I've played around with manual vs. auto and internal vs. external using the Boson App. Manual/external does produce a nicer image, but the system software still tries to command a flat field every so often. Even with the Boson shutter disabled and in External mode, when the system decides it needs a NUC, you need to quickly install the lens cap, use the palm of your hand, or provide the camera with some other uniform source before the NUC fires. Wish FLIR would change the PTS software so NUC is only controlled by the Boson. Quality of the source has a lot to do with the quality of the post-NUC image.
Auto vs. Manual NUC should be a user preference. Some folks don't want to be hassled with the operation, others want to control it. Think it depends on the user and the situation. I think the Trijicon products do a nice job with their NUC process. Identifying the best target is the most difficult step.