Earlier quoted context omitted.
You're focusing on the specific example, rather than the general idea that your purchase history indicates your riskiness, and an insurance company who has that data available to them will adjust your policy based on your behavior. And why shouldn't they? It's the same concept as you indicating whether or not you smoke, how much you drive to/from work daily, how many driving tickets you have, etc. Questions you willf…
The flip side of that is that people who don't provide that information are going to be grouped into a pool of relatively high risk people and be forced to pay higher bills. And the incentives line up such that, if you're better than the average non-sharer of personal information, you'll be significantly better off switching to being a sharer, which in turn increases the risk profile of the remaining non-sharers. And…
So why not use examples like these? "Oh, you bike to work? Great--your bill is going down! Oh, you have a good credit score? Okay, your premiums just went down again. Oh, you eat salad once in a while? Great, it doesn't take much to eat more healthily than the average American, so--you guessed it--your bill just went down again!"
Less information asymmetry between customers and insurance companies would lead to better prices for safer people, and fairer prices for riskier people (by "fairer" I mean "closer to their average lifetime insurance payout).
Whether it's worth it in terms of the loss of privacy is another question, but we shouldn't stack the deck by pretending everyone's premiums will go up unfairly.