>never has prevented fraud.

Interesting, I've heard otherwise but it was anecdotes. Do you have any data on that?

> to track non-fraudulent users

You listed a large number of ways to fake the phone number which is why you believe it doesn't prevent fraud. What is to stop a non-fraudulent user from doing the same thing to prevent the tracking by the company?

>Do you have any data on that?

The original stated intention of the practice was that "it" [mandatory phone number registration] "prevents fraud" (though this stance was being critiqued by the person who raised it, not defended).

I'll concede that it probably has stymied some of the most trivial, incompetent fraud attempts made, and possibly reduced a negligible amount of actual fraud, but the idea that it can "prevent" fraud (implying true deterministic blocking, rather than delaying or frustrating) is refutable by the very reasonable assumption that there is almost certainly no company that implements mandatory phone number registration that has or will experience ZERO losses to fraud.

That said, in fairness, this is an unfalsifiable and unverifiable claim, as to my knowledge, there is nothing resembling a public directory of fraud losses experienced by businesses, and there is no incentive for businesses to admit to fraud losses publicly (they may have tax incentives to report it to the IRS, legal incentives to report it to law enforcement, and publicly traded companies may have regulatory incentives to at least indirectly acknowledge operating losses incurred due to fraud in financial reporting), but that doesn't make the claim itself unreasonable or improbable.

>What is to stop a non-fraudulent user from doing the same thing to prevent the tracking by the company?

The argument isn't that mandatory phone registration unavoidably forces privacy infringement upon all users, just that it does infringe upon the privacy of some (I'd suggest a vast majority) of users in practice.