You lost me on this one. If the logs show the go pedal was pressed why isn't that proof?
Maybe sensors was faulty and reporting that the pedal was pressed. Logs would only show what sensors have detected.
Or maybe the algorithm that is calculating the pressure has a bug, and it causes misinterpretation of sensor output.
(Or maybe there was a physical fault.)
Of course, there is no way for us to know, all I'm saying that these are possible.
Toyota has history of unintended accelerations. They always blamed users first. Besides from sticky gas pedals and bad floor mats, they have a case that they couldn't find the problem and blamed user of pressing the gas pedal instead of break. But further investigation showed that Toyota's code base is a mess and has a bug in their car's software. These slides is a nice: summary
https://users.ece.cmu.edu/~koopman/pubs/koopman14_toyota_ua_slides.pdf
Thinking about Tesla cars' bugs and software complexity of the system, in addition to engineers who are working overtime in a stressful environment, I may as well guess that Tesla's code quality is not good.
Elon Musk stated a few weeks ago that they are working 7 days a week. They've found a bug just a few days before the release and tried to fix it in a few days, during holiday period! Sorry, but an engineering team that is doing overtime all day time and working during holiday period can cause lots of bugs. It's not a good thing, people shouldn't applause Elon Musk, they should be afraid since Tesla is rushing such an important piece of software that may cause lives.