What's interesting about this is that it
appears they did! Very very early SB models exist and tested MUCH better than the first run of production hardware. You can find a lot of early reviews that praise the battery life, etc.
Then the first wave of consumer-facing SBs went out and it was a total disaster. This might be something that Microsoft can fix, because they have dramatically improved the product experience and been very receptive to trading defective hardware. Mine was traded up the instant they looked at it, with apologies and a customer care call.
> they should have tested Skylake, and it in the Surface Pro 4 prototypes, detected the problem, and delayed the launch. Intel's problem became their's when they shipped it in their hardware.
Did you say the same thing when Apple shipped a massive defect rate on their first gen retina macbooks? Because they DID test thouse, and they still ended up shipping a truly phenomenal number of lemon screens with huge defect and failure rates.
Oh, and Apple refused to replace all but the most egregious failures. I still have a machine with such significant ghosting that it can be difficult to use. Ironically, DaringFireball is actually unreadable. I keep this machine around because it was part of a very special segment of my life, but also because I like showing people, "Yes even Apple's legendary hardware is rife with first gen bugs, and your iPhones and hypothetical new macbooks are no different."
