There are low-latency USB audio interfaces that do better than 10-15ms, and it has more than enough throughput for 40 audio channels, although obviously the latency is still not as low as a PCIe device over thunderbolt can go.

While I agree that thunderbolt is primarily being adopted by the professional or pro-sumer market (the only thunderbolt devices that I've ever used apart from some of Apple's very affordable gigabit ethernet adapter is pro video gear), I think that's primarily because Intel is charging too much for the controllers going into the cables and devices.

Belkin's "doodad" costs $200, which is insanely expensive if you want a USB 3 adapter.

TB2 has the same aggregate throughput of TB1, it just combines the existing channels. That's great if you're only connecting one device, but not so great if you're daisy chaining. TB1/2 don't lack power, they supply 10 watts, far more than the cable itself requires. That's already double what USB provides over data connections, and you shouldn't be drawing much more than that from a notebook anyhow. The vast majority of systems sold with TB support today are notebooks (mainly because the vast majority are Macs, which are overwhelmingly notebooks).

That's not native. The new version of Thunderbolt does it natively.

Not unless they want to be stuck on the current version of Thunderbolt.

Corning owns Gorilla Glass, though, so that wouldn't help Apple increase their vertical integration. TFA does mention that there is work in that direction going on, though.

Errm, sapphire *IS* transparent aluminum. It's aluminum oxide (Al2O3).

From TFA, Gorilla Glass reportedly costs about $10 per display when Apple first started using it (on the original iPhone), with prices eventually dropping to about $3 today. Sapphire is expected to cost about $20 after all is said and done. It'll come down, certainly, but it's definitely not cheaper.

As for being tougher, my understanding is that it's far more scratch resistant than gorilla glass, but not necessarily as shatter resistant.

And yet they're adding support for running USB 3.0 over Thunderbolt with the new standard; this new standard was the perfect opportunity to unify the two standards, but they stopped short of supporting the latest version of USB or using a compatible connector.

What does Apple have to do with it? Intel manages Thunderbolt, not Apple.

Is Intel *TRYING* to kill off Thunderbolt? They can't make up their mind if they want USB 3.X or Thunderbolt to be their next-gen connection, and now (despite extremely low Thunderbolt adoption), they're going to change the connector?

USB 3.1 and Thunderbolt are redundant. At this point, they even both support uncompressed video. Pick one, drop (or deprecate) support for the other, and the industry will migrate.

Tesla has plans to add automated battery-swap stations to some of their superchargers. They've already demonstrated the swaps in a controlled environment (they put a battery swap station under a stage, drove a car on stage, swapped the battery, drove it off, drove another on, swapped it, drove it off), although it may have just been taking the battery off and putting the same one back on. All sedans they've sold to date are designed for battery swaps, and they plan to cost it out as an equivalent to what gasoline costs in the local market.

In terms of reducing peak loads on the grid, they do have plans to combine grid storage with solar panels (how convenient, Musk has SolarCity building solar deployments and Tesla building battery factories), with the goal of having the superchargers produce more power than they consume.

Tesla also plans to have cars in multiple price ranges; it sounds like they plan to have three within the next few years (the Model S, the Model E, and the Model X, and yes I realize the pun they built into the model names, they also trademarked "Model Y"). One of them, the Model E, is going to target that $30k price point (at least after rebates), although the reports are that the range will be in the 200-250 mile range rather than the 300-350 that their current top-end cars get.

In terms of increasing the range, that should improve gradually over time as the cost of battery cells go down. That was the point of the $5 billion dollar GigaFactory that they're building, to reduce the cost of lithium ion cells. The primary goal of that is to reduce the cost enough to hit the Model E's price target, but it also has the benefit of enabling higher ranges in the luxury cars where they can spend more money on the battery, albeit at the expense of weight. I know that they're working on longer-term solutions to improve range. They got some patents recently that relate to combining metal air batteries with lithium ion batteries in a hybrid power solution designed to circumvent some of the limitations of metal air batteries (they have the potential for higher densities, but have poor cycle life), although that stuff is a rather long way off.

In terms of ubiquitous fueling stops, they're working in that direction. They're hitting a steady pace building new stations, and by the end of 2015 should have most use cases covered between home-charging overnight and superchargers for distance drives. Automated battery swaps may help too.

Their success isn't a sure thing, but they're definitely making progress towards solving the problems.

Musk has personally guaranteed the resale value of the Model S against any comparable luxury sedan. The "personally" part meaning that if Tesla goes out of business, he will still honour that guarantee. I presume that means if you sell it and are unable to get the guaranteed value, he'll make up the difference. As a result, if Tesla were to go bankrupt, you could avoid the liability by immediately selling your Tesla car and relying on the guarantee to avoid losing money on the deal. Of course, if ALL of his companies go under, you could still be in a fix, but this isn't 2008, and he's not strapped for cash.

Some of them are sort of valid, but not relevant in practice. For example, it's true that Tesla's current service infrastructure can't handle high demands... but that's because the infrastructure is sized for the current customer base. Building a service infrastructure that can handle many more customers than you actually have is a waste of money, and it's completely unnecessary so long as you continue to scale that infrastructure as you grow.