I can't find the numbers easily right now, but I'd really like to see "deaths / person-hours flown on a 737 Max" vs. "deaths / person-hours driven".
The U.S. motor vehicle accident rate is about 1.2 fatalities per 100,000,000 vehicle miles travelled. Figure a car's average speed is 30mph so 1.2 fatalities per 3.3 million hours = 0.36 fatalities per million hours. Therefore on an hourly basis, the 737MAX is very roughly 5x more dangerous than driving. OTOH the 737MAX averages about 400mph, so per mile it's probably something like 2-3x safer.
If you’re over 10K feet in the air there is some room for recovery procedures, but if you haven’t reached cruising altitude you have much less time before a plane finds itself impacting the ground.
Per-mile travelled isn’t an adequate standard to compare aircraft with other modes of transit when talking about safety.
Besides, when people describe flying as "safer than driving", the implication is that you're more likely to end up dying one way than the other. So any measure of "deaths per x" (x is miles, hours, or whatever) needs to be followed with "typical x per lifetime".
Scrolling with my mouse is super safe – a system that can crash you into the ground at 500+mph and that relies on single (possibly faulty) sensor is not super safe.
Current aircraft and flying is very safe. Even with the recent 737 fiascos, flying those is by far safer than driving to work. If I am OK with taking a road trip on vacation I am OK with flying Max.
This is not to say that improvements, re-certification and additional pilot training are bad. But from a purely practical point of view (which I try to practice for routine decisions), if we want to improve safety, there are better areas to spend $$ and / or hours than on endless re-polishing of a system that is super safe already. Life is dangerous; estimate your micro-deaths and extract most utility and fun from each of our 1e6 microlifes. My 2c.