(iii) By Tesla's own admission, systems of the car were still "in Beta", a status that (a) is very unusual for a passenger car, (b)
makes the car different from virtually all other cars on the road; and (c) essentially admits that the car deserves more intense scrutiny than cars that use proven technology.
In terms of scrutiny it's a byproduct of it's popularity. It's a little hard for Tesla to go under the radar when there is a Tesla Roadster out there in space orbiting around the sun.
From the NTSB's perspective it's a pretty straight forwards investigation, and is extremely important because EV are getting more popular. They absolutely have to investigate it as a way of learning the best ways to handle an EV accident. In fact I would say the investigation itself is proof that Tesla won.
In terms of beta I'm afraid that if you want a high tech car with OTA updates that it's always going to be in some form of beta whether a manufacture uses that terminology or not.
Subaru's initial eyesight system had issues with false braking due to shadows. They never called their system as being in beta. Clearly it was despite all the validation testing.
Volvo improved their own system through a software update to detect large animals. They evolved it, and I imagine they'll make further tweaks to it.
None of these technology is fool-proof. That's why they need to evolve, and change.
In fact MB claims they dumbed down their system to make it less likely that a driver would grow too comfortable with it. Obviously it's pretty easy to release something as not beta when you have such a small goal on the deliverables.
The only company I'm even aware of that REALLY went through a lot of work to get their system solid was GM with the SuperCruise system. They delayed it well over a year for validation/verification. What you get is a system that works great over white listed roads. The reason it can work so well is because of the limitations placed on where it can operate. It's a truly hands free system so it makes total sense for it to work the way it does.
Autopilot is really hard for a lot of people to wrap their heads around. It is because it's not a system set in stone.
It's this evolving thing that changes over time.
You pay for it before it is ever done, and it's always late. Heck Tesla might as well sell it on Kickstarter.
Things are added to the neural net it uses for vision detection (lines, cars, pedestrians. etc)
Things are subtracted out
Algorithms are tweaked
Massive amounts of data is analyzed
Critical systems such as AEB are validated using Shadow mode on the entire fleet. This reduces the chance of a false positive because the car can compare it's desire to do an emergency stop with what the driver did.
It's constantly loading AP tiles, but doesn't tell you anything about the status of them.
The hardware changes and then you find yourself looking for posts verygreen to see what changed, and he doesn't even have a Tesla anymore (apparently not just the software was beta on his car).
It's is extremely controversial because two statements are true at the same time
Releasing AP2 among the masses in such an alpha form like it used to be (before this fatality) was completely irresponsible.
AP2 will reach a much higher level of driver safety than it would have achieved had Tesla only tested it in house. Of course it was because they made their customers the testers.
To Tesla the most shameful outcome of this accident isn't if this driver was using AP, but if the driver wasn't. If it was just another accident that happened due to driver inattention/distraction. I generally dislike lane-keeping/steering/departure-warning systems (for lots of reasons), but even I have to admit they save lives.
When I drive my moms Subaru I turn off the lane departure system, and I never use the lane-keeping ping-pong assistance thing. I do because their both lame.
When I first got my Tesla I turned off the lane-departure warning because it was pretty lame.
The closest thing to actually liking/using aside from TACC was AP. This is extremely important because to save lives people have to use the system. I hope by the time Tesla makes it to a full EAP system that it will be good enough for me to use it.