Moore's Law is shining through here. Though Moore's Law refers to technology changing in 18 months for no price difference. This is probably going to be a 72-month gap if it launches November 2017. That's 18 * 4 = 72. Technology should be eight times greater if technology doubled four times, without changing costs, since the launch of the original Xbox One.
If we compared the Teraflops Scorpio promises, 1.31 Teraflops was the inital Xbox One amount. If they are targeting 6 TF, then Scorpio is about 4.6 times greater than the Xbox One, meaning Microsoft is going to be making a profit on these machines. I'm not saying teraflops directly correlates to costs, but it's my assumption under Moore's Law that it isn't exactly the strongest computational power per cost that Microsoft could have put in for $399, or even $499. Which is fine, as they made huge losses on the original Xbox One by selling as low as they did, removing the Kinects, and having to pivot their core competencies, while also having to support the original console for a while is going to be taking a toll on the Scorpio's potential and on Microsoft's wallet.
Do I believe Moore's Law is always accurate? Nope. There's always something that will create a roadblock, but it may explain the reasoning of the Xbox One and Scorpio as it is.
