What incentive does a tier-1 provider have to collect data from their clients? Keep in mind their customers are not individuals, but large orgs: data-centers, corporations, office-buildings, universities, et al. They lack the strong correlation of "1 IP" => "1 customer" that a residential ISP tends to enjoy; they also lack the storage infrastructure that an AMZN, GOOG, MSFT, et al. tend to have as a function of doing business. (Since the problem tier 1 providers aim to solve tends to require large workingsets, not large datasets at rest per se, they tend to need RAM more than durable storage.) I imagine the infrastructure needed upstream to track such traffic would be immense & costly, and more importantly it would just not be very applicable to the provider's major product offering.
Furthermore just what would their end game be? Per all the DOCSIS whitepapers I've read: my residential ISP intends to sell me any number of "over the top" services: a plethora of cable channels, their own streaming services, VoIP, alarm systems, whole home DVR, etc. There is a lot of money to be made there in terms of equipment rental, upkeep, and paid programming. More importantly most of it goes right into their pockets. The way I see it, it's not about selling the data, it's about using it themselves.
Compare that to a tier-1 provider who has one job: get a drop to my network fabric. Their business revolves around (A) doing that regionally, (B) maintaining good peering, and (C) being extremely competent network engineers. As I see it a tier 1 provider has far less incentive to spy on their users compared to a residential ISP. This doesn't obviate the need for caution of course, since nation-state level actors have all the more reason to spy on tier-1 providers simply due to the volume of traffic that can be intercepted.