That statistic is not controlling for the drivers. It's interesting, but it may be biased by who buys Tesla and how they drive. Similar to likely "cars with aggressive custom paint jobs have higher fatality rate".
Maybe. Taxi drivers have different incentives/experience/routes than the other population. It may apply to them, or not. For ride share, the choice of Tesla may have been mainly economical for them.