I don't think any of these people who have dealt with "unintended acceleration" will ever believe they did anything wrong until someone comes up with a device that mounts under the steering wheel, and video records feet/pedals.
I do take issue with the "...having reviewed the logs and that it shows that the accelerator (throttle) pedal was pressed during the accident." The logs don't show that. The logs do not show the ACTUAL position of the throttle pedal; they show the ELECTRONIC position of the accelerator pedal position sensor. That sensor is basically a potentiometer (a variable resistor). Anyone who has dealt with electronics has dealt with faulty pots before. They fail in one of three ways; they get "dirty", they fail open (0%), or they fail short (100%). If the pot failed "short", it would look (to the logs) like the throttle pedal was pressed to 100% (that she floored it).
Now, car companies that use electronic throttle position sensors normally have redundant pots built in, with logic in the system designed so that if one pot gives a bad reading, it gets ignored (or is given a lower priority). i.e. if one pot shows a throttle position of 100% and the other shows 15%, the car assumes the 100% reading is faulty. I can't/won't speak for Tesla's circuit design, but would be shocked if it was designed with a single-point-of-failure (especially something like a throttle position sensor, with is a moving part and will, by definition, eventually wear out).
Now, I'm not saying this happened here. I'm willing to bet some money that she pressed the wrong pedal.
I couldn't help notice that, in the article, the car owner (the husband) went on and on about how familiar he was with Teslas ("I have owned several Model S Tesla’s and currently own 1 P85D and 1 P90D"...yada, yada, yada). That's great, that's wonderful, you're an expert Tesla driver. However, YOU WEREN'T DRIVING.
Absolutely. Yes, the cause of unintentional acceleration is normally driver error.
However, there is some debate as to whether the unintentional acceleration in the Toyotas was a coding error or not. All we know is there were coding errors found in court, and Toyota paid a settlement, and made some changes. Whether the mistakes caused the crashes is not known.
A coding error is not going to show up in a datalog. The odds are slim that the problem lies in the coding, but thinking it's impossible is not correct.