You and a handful of other posters use this same tactic when people questioned nVidia when they began to release their big chips as "Titans" only to lower the price on an artificial 6-month cadence. I have a right to discuss what I believe offers value in consumer products. You're answer does nothing to validate any type of opinion for or against the topic at hand.
You have a right to discuss what you want.
Which makes the use-case almost non-existent, as this price increase justification (while still a very weak justification) is now almost purely for people that game.. yet the discussion has been that this many cores is a bad choice for gamers..
Let's drop the pretenses, for most games, a hyper-fast quad core is still going to smoke these high core count chips from both Intel and AMD and it'll require much less exotic cooling, too.
Skylake LCC is a 12-core die. The $1000 10 core chip has 2 cores already disabled.. not perfect.
LCC die is 10 cores. MCC is 18 core, HCC is 28 core.
Four defective cores. And I have to stop you there for a second. This is Intel's three year old 14nm process. Skylake's core itself has been in manufacturing for over two years itself. You have praised Intel's fabbing prowess! Skylake-X LCC looks to be about 280mm2.
Two defective cores. Also, the 14nm process being used to build Sky-X is not the same 14nm process that was used to build Broadwell-E, Broadwell-U/Y, or Skylake-S.
Further, the Skylake core inside of the SKX chip is actually quite a bit different. L3 cache structure is totally different, L2 cache is much larger (1MB vs 256KB), the CPU core has AVX-512 support (and this is a pretty significant addition in itself).
There is zero reality reflected in Intel's pricing. Intel, supposedly the best fab in the world, who is making 280mm2 chips (bit larger than Polaris) on a three year old process is having bad yields on fully enabled chips? C'mon on now don't make it that easy...
Die size isn't all that matters, your clock speed/power consumption targets matter, too. Selling parts in which all 10 cores are guaranteed to run at 4.3GHz (and two guaranteed to run at 4.5GHz) is no easy feat, especially with the much larger L2 cache in this core vs the client SKL core.