Actually, NVidia have been doing the same thing for a couple of generations of GPUs as far as anyone's been able to tell, the press are just a lot less willing to kick up a fuss about anything they do than with AMD. (And I mean literally the same thing - designing their cards so the actual clock speeds and performance they run at out the box varies from card to card, then cherry-picking the best ones to send to reviewers.)
Lightning doesn't support USB 3.0 or proper video out, and improving it to support them would be more work than just creating a new, incompatible connector if it's even possible to do so at all. (Lightning doesn't have enough pins to do USB 3.0 without some kind of intermediate translator chip in the cable, and as far as we know the highest-bandwidth protocol it currently supports is USB 2.0 so you'd have to somehow create a new intermediate protocol too.)
Apple may never have claimed the cable restrictions had anything to do with safety, but it's easy not to realise that due to all the Apple fanbois in the tech media and on
There is no silicon between the two devices if you're using USB over the Lightning connector, that's the point - early on in its life, someone tore the cables apart and the data lines are wired straight through, the authentication chip can only communicate with the iPhone and then only at speeds slower than USB 1.1 Low Speed.
Trust me, the old SCART cables that were used for the same purpose in Europe in the standard-definition era were worse - they were just as orientation specific and flimsy, except they were also much larger and the cables attached to them much heavier.
Lighting's also just a connector for USB 2/3 - the data wires in the Lightning-to-USB cable go straight through. The reason the cables are so expensive is because they have to have a special Apple-supplied lockout chip or iPhones and iPads will refuse to work with them.
Probably still better than Intel's Galileo board, which doesn't even have proper native GPIOs (they all go through a slow I2C I/O expander), is more expensive, and has worse power usage.
Did you actually bother to click on the second link, written yesterday, which is all about how the problem is still there even after the supposed fixes? Be sure to read the second page too.
Tesla advertised their car to owners as actually having the much lower levels of power usage it would have if the standby functionality was working corrrectly. They completely neglected to mention that it wasn't functioning and that the car used far more power when not in use than they were claiming. It's no different from advertising a TV as having ultra-low standby power usage when in fact the manufacturer knew it would draw far more power.
It almost certainly just acts as a transparent proxy that intercepts connections and DNS requests and sends them through Tor - there's already support in the Tor client for doing this.
Of course, your example shows one reason why any statement about the Model S's safety from Elon Musk should be taken with a pinch of salt - it's just too new! Capacitors with that issue generally took well over a year to go pop (that's partly why the capacitor manufacturers didn't cotton onto the problem), so if the Model S had exactly the same fire issue you were complaining about Elon Musk could still portray it as less likely to have a fire because it hasn't existed long enough for that kind of problem to show up.
To be fair, if that had happened to Model S's then Elon Musk and his supporters would be spinning it as a good thing and the media coverage as some kind of anti-electric-car crusade just as he is with the Model S fires - after all, not only did no-one get hurt by the fires, there was essentially no chance of anyone getting hurt because no-one was in the cars at the time, and it only happened as a result of them being underwater.
You're also teaching them not to recognize when calculations make no sense. How can you subtract 5 pennies from a cup of 6 units of coffee? You can't, and that kind of check will be important in a few years once they move onto using that maths for real-world calculations where dimensional consistency is important.
If you read the sentence before that: As single bits in memory control each task, corruption due to HW or SW faults will suspend needed tasks or start unwanted ones. It only took a single bit in non-error-detecting RAM getting flipped to cause that particular fault, something that could easily happen due to cosmic rays or minor radioactive contamination in the ECU packaging - and that's before you even take into account all the other potentially memory-trashing code. It's more like a manufacturer deciding not to ground the case at all and just hoping nothing will come loose and short to it.
They found, amongst other things, that single-bit flips in non-error-detecting RAM could cause unintended acceleration. Those aren't exactly uncommon and can be made even more common by things they didn't investigate like the materials used to encapsulate the chips.