As a PC gamer in the 90's....
PC gaming wasn't what I would call "low quality" back then, in-fact with my rose tinted glasses on, would consider it to be the golden era, there was almost half a dozen PC exclusives launched every month!

I thought the golden era of PC gaming is NOW! PC gaming has never being more healthy than it was before with it eclisping console revenues.

With titles such as Master of Orion, Dune, StarCraft, Diablo, WarCraft, Settlers, Sacrifice, Evolva, Battlezone, Age of Empires, Freelancer, Freescape, Starlancer, Wing Commander, Arcanum, Icewind Dale, Neverwinter Nights, Baulders Gate, Battlefield, Black and White, Anno, Mechwarrior, Caseser, Chrome, Civilization, Alpha Centauri, Comache, Command and Conquer, Dark Reign, Total Annihialation, Elder Scrolls, Deadlock, Descent, Quake, Unreal Tournament, Doom etc' etc'.

Starcraft was an especially worthwhile title back in the day when PC was getting overshadowed by consoles. 

And if you ran with a 3dfx voodoo, you had Glide, you could run at 1024x768 resolution with 2x Voodoo 2's in SLI and you had the best-in-class image quality and performance when most console games were 640x480 or lower and only 30fps, in-fact back then you could run at a resolution equal or greater than some Xbox 360/Playstation 3 games!

These days... It's all crappy lazy console ports with very little time or effort.

As for SteamOS, I learn't not to doubt what Valve could do, when steam launched I laughed and thought it would fall flat on it's face... 75 million users later...
So, I'll take a wait-and-see approach.

I would also like to wait and see but this thing will literally crash and burn so hard like no other platforms before it ...

Up until about 99... sure?

Late 90's sure.  A drop didn't happen until 2000.   Coincidentally around the time MMORPG's starting blowing up, making retail less important.

So this means that PC has declined sooner than I thought. 

Also right around the time digital started being seen as a viable market. (Though small at the time).   

Steam debuted all the way back in 2003.

PC Retail continues to drop... as retail becomes more and more irrelevent.  

Gameplay quality wise... the best games on PC aren't on consoles if you ask me... and as I imagine most PC gamers would agree, based on revenue generation of big franchises.

You are right about the best PC games not being on consoles but they continue to keep sharing a large amount of high quality AAA titles. 

If we use your metric of money... League of Legends, WoW, Dota 2, Civilization.  The big PC revenue makes COD look lame by comparison. 

Which your revenue based arguement seems to argue is the pinacle of console gaming.

It is true that PC gaming is getting better revenues but PC gaming was quality all the time when consoles in the past were known for their better revenues. 

That's not even getting into the casual juggernauts.

It's not Windows... or OSX.

The best part about Steam OS is that it built from the ground as a gaming OS, and not as a multi-purpose OS with the gaming bit added as an extension up over the years.

But do see clean game optimisation ? Like we are winning real potentiel or its still pure theory :o ? like going to High to Ultra like ?

 I Am wondering If I will need it as absolute in a part of my drive :o ?

or its more like a Android - apple kind of stuff matter of taste ?

...Really ? I wasn't expecting samsung's 14nm process node to match Intel's 14nm process node but I was assured that it's chip area scaling was 15% better than TSMC's 20nm or 16nm process node so relative to Intel's 14nm process node the size of samsung's transistors would be roughly 18.5nm so to speak. The reason why global foundries went to the trouble of licensing samsung's 14nm process node over their own 14nm process technology was because they could meet their customers demand on schedule while being able to produce chips at a marginally smaller size. This would literally put globalfoundries ahead of TSMC in semiconductor manufactoring technology by roughly 12 months due to the fact that they will have implemented FinFet technology first while having better chip area scaling too! Mass production is slated to be around early 2015 for global foundries.

Steam Machines will have a hard time for sure, they're kind of an inbetween product that are neither here nor there, in my opinion and I haven't really seen much incentive for me to want one yet, and that is coming from possibly the biggest Valve fan in the Northern hemisphere.
My initial and main concern is the pricing though; they seem to place themselves above the One and PS4 in price on average, you get far more powerful hardware but that is certainly not a guarantee for sales (see original xbox for an example) and the somewhat wonky gamepad will likely turn a few people off, that also symbolizes the inbetween nature of the product for me; it ends up somewhere between traditional and touch and might very well miss both.

Gabe Newell and Valve; I love you and everything you do to bits and I consider this company to be the most positive influence on gaming in general in the past 20 years, along with the likes of CDProjekt, but I think you're making a big mistake with these machines, I simply cannot picture a sizeable market for them.

How is execution the issue when it was just a bad idea from the gecko ?

Even if steamOS had everything except for directx it would still fail miserably because it offers no differentiation between the PC and the console market. As blurred as the line is between console and PC they atleast have a factor that differentiates them such as easier accessiblity as well as having some high quality exclusive software. PC's have benefit of being able to do many things not related to gaming. Steam machine on the other hand will have almost nothing to differentiate it because Valve can't afford to support a huge platform as large as this. Valve has also failed to control the price that OEM's are charging which will mean that it will get a bad reputation as an overpriced console like the PS3 initially was. The controller is shit too as well.

Why didn't they realize all this ? 

The plan was bad all along and I just can't see it having any sort of potential.

- From what I've seen elsewhere 15% scaling if you don't want ANY power or performance increase with that shrink. Also, it's compared to "a 20nm planar process" which doesn't actually exist. So the numbers are meaningless.

- Do NOT believe that timeline for Globalfoundries. Trust me, no production until 2016, definitely after TSMC.