>Well if noone who owns these devices wants to do it it probably shouldn't be done then, right?
I think it's just a matter of comparing the number of people who would actually opt in to such a service versus the number of people who don't really care one way or the other whether they're opted in.
When it's opt-in, you can expect your network to consist of:
total = [number of people who care enough to opt in]
When it's opt-out, you can expect:
total = [number of people who care enough to opt in] + [number of people who don't care either way] + [number of people who who would opt out if they knew they could] - [number of people who opt out]
Even if you perfected your user education to ensure 100% of [number of people who who would opt out if they knew they could] actually opt out (a perfectly-valid "moral" end-goal, IMO), your resulting network would still be significantly larger than if you made the feature opt-in.
If a decision-maker sees the feature as "helpful" to users (a likely assumption in a vacuum), I could also see them making a judgement call to make the feature opt-out to maximize the "good impact" it potentially could have.
But this is Amazon. So it's way more likely the decision-maker is just trying to maximize company profit (which could still be morally justified if they cared to ensure everyone that wanted to opt out could!) and not thinking about whether the feature morally "should" be done or not.
I'd give any other company the benefit of the doubt here and assume they're valuing the good of [total number of people on the network] over the "bad" from screwing over [number of people who who would opt out if they knew they could]. I'm inclined to think Amazon is valuing the profits of [total number of people on the network] instead.