Th studies in your linked are all bunks, they do not reflect real life condition because they were all simulated.
LOL, and yeah they are going to set up "real life" experiments and kill a few people to prove exactly the same thing ?
Yeah right.
It is common knowledge that if you have an operator of machinery, factories, heavy equipment (and yes cars) with his brain underutilised actually doing the task, their reactions when they need to respond will be slower...it's science, it's designed into factory workstations and everything.
If you are on cruise control, you've travelled an extra some number of feet at CC speed before you touch the brake...if you aren't your foot is closer to the brake (on the accelerator), and the mere fact that you've lifted off to get to the brake means that you've saved some speed...surely you can grasp even THAT simple point.
It's science, if you take away inputs from the operator, their mind will wander, and they become inatentive, and it takes longer to shake them back into the here and now.
Tesla need to either take control entirely, and the responsibility that comes with it, but "beta" testing on public roads with a disclaimer that "you must be in a position to take control if needed" isn't providing that level of protection...clearly.