A reason alluded to, but not directly stated or explored, is that car insurance is a low margin / high competition business. That 1% underwriting profit is a good indicator of that - actual profits come from investment of the float (and understanding this concept in more detail is a good reason to be reading Buffet's Annual Letter to Shareholders - I've been reading them for a couple of decades now, and they're all available on the Berkshire investors site).
The problem with a 1% underwriting profit of course is that is an aggregate value over the industry, over the years. Some years are better, some are worse.
That thin margin is also our protection as car owners. An insurance company that is badly misplacing Tesla car insurance to the high side will earn some incremental underwriting profit, and will lose business to other companies that aren't.
One of Geico's strategies, talked about 1 year by Buffett, was to make their premiums easy to see, and to keep them low (which puts heavy emphasis on their ability to price correctly). The way they see it, it's a win either way. Showing lower premiums will bring in customers. Accurate pricing of those contracts at low prices, where the customer goes elsewhere even cheaper, will be losing contracts for their competition.
That doesn't mean that Tesla can't develop a lower priced insurance using superior information at the individual consumer level and their individual driving habits. Only that I don't see it as a no-brainer (mostly because the loss of privacy will be an impediment for many potential consumers).
Then again, I think I'm personally closer to $1200/year on our Model X insurance with State Farm, partially due to having our entire collection of insurance coverage with State Farm. So maybe I'm just happy with the rates we're getting, and don't feel a need to go looking for something better.
