So I hate logical positivism. From logical positivism we get the philosophical foundation of modern tracking and advertising, that every data point obtained about a subject is necessarily meaningful and actionable, and that which cannot be measured does not exist. From there we get the unhealthy Nielsen-style reductionist tendencies in marketing that believes all it needs to know about you is your age, race, gender, sexuality, and income bracket. I think Frankl's discussion on dimensional ontology is the most complete picture I've found of why this is a flawed point of view, though I think any person who has been discriminated against based solely on one of those criteria should be able to understand the limitations of that model.
I hope we can start to understand better as a tech culture that, unfortunately, human beings are decidedly unscientific in the sense that past data collected will not always provide accurate conclusions about future behavior. People grow and change, and collecting data silently in the background and trying to draw actionable conclusions without active user input and feedback is a fundamentally inefficient process, with harms that can be quite difficult to measure, though this article articulates some of them quite well.
I suppose I can imagine that if we lived in a society free from the incentives generating the kind of abuses this article lists, and someone has generated an algorithm that could help you find meaning and success and fulfillment and loving relationships based on all the tracking data you provide, I can imagine this to be the dream of big data for which the sacrifice of privacy might be a necessary one. But given that we still live in a society that prefers exploitation to altruism, what we can salvage of our privacy is still probably the best option until we can reach that next stage of our evolution.