So long as the methods of the study are fully-disclosed there is no issue here, and no invasion of privacy. You would prefer not to participate in such a study

So long as the methods of the study are fully-disclosed there is no issue here, and no invasion of privacy. You would prefer not to participate in such a study, and that is perfectly fair. That doesn't mean its an ineffective method for observing how users interact with software. I've been surprised many times with (in-person) usability studies where we ask users to perform certain tasks with the tool. It's a great way to improve UX.