There are regulations in place that would prevent Tesla from suddenly becoming the country's greatest and most profitable insurance company. If you don't believe regulations exist on insurance companies profitability, there's really no need for further discussion.
Ok, here goes. In the US there are no material Federal regulations regulating insurance companies, apart from general Federal regulation of public company actions such as from the SEC. Insurance regulation is almost exclusively at the State level and some States work hard to make regulations quite vague. In fact there are often regulations defining what insurance is, and most States establish some sort of payout ratios, although this are quite easy to 'manage'. In sum, in the USA insurance definitely is NOT WYSIWYG. There are endless examples of legal but nonetheless popular 'insurance' products.
In order to adequately explain this we would go seriously off topic so I will not do that.
Tesla approaches to insurance as they have been explained in the past, will be more data-intensive than have been previous auto and homeowner products.
Soem commercial insurance companies, like ProgressiveSnapshot, are beginning to directly measure actual driver behavior and set rates accordingly. Those practices have been increasingly common in fleet coverages as well as some aviation and marine categories. All share actual remote performance data, as such they are all quite invasive. Tesla already has much performance data but has not yet asked for or directly applied such data in rate setting or any other such case.
Tesla certainly will not try to become "the most profitable" insurance company, but maybe they will become "the greatest". Under most State rules there are no existing prohibitions of such underwriting practices. Existing actuarial techniques need not even change, just the data and the continuous revisions would require new business approaches.
Further discussion is needed, especially for people who do not understand how Tesla views the evolution of their insurance products. Their views have been open, but thus far nobody has been asking many questions because they've not yet done anything.
Since I personally have been following Elon Musk: 1. PayPal could not be successful, 2. SpaceX was a joke and would surely destroy Mr. Musk, 3. Tesla was stupid and ridiculous and will fail just like all the others. 4. Tesla did not know how to build a car, 5. Starlink will never ever happen, 6. The Boring Company will never actually boil anything.
Now some people think Elon knows nothing about insurance. By now they should be wary of assuming facts not in evidence.
As for subscription models for things like FSD and much more. By now they should be wary of assuming facts not in evidence.
Finally, fact in evidence is indisputable. Elon Musk frequently underestimates how long it will take to solve some particular problem. Many of us still think it is a failure when something takes longer than it has been forecast to take. That should surprise none of us. The wait will be irrelevant when the thing actually arrives.
I've waited years for Tesla three times. I'll keep doing it. It's worth the wait.