It's a classical case of diminishing returns.

In example going from a "cylinder" with 6 polygons in the walls to one with 60 is a much bigger visual improvement than going from a cylinder with 60 polygons in the walls to one with 600.

Also I firmly believe that the games with "realistic" graphics will continue to be firmly rooted in the uncanny valley for the forseeable future.

I rhather see the console makers spending the uppgrades on specialised AI hardware than spend more on graphics hardware. There are more things to games than looks.

The modern games I've played (mainly on PS4 and PC) still are lacking in the interaction aspect with NPCs/enemies, even if refinements have been made the scripted nature of the AI still prevails.

This comparisons redundant these days, because graphical upgrades are rarely about polycount and more about scene complexity, shaders, lighting, etc  the polycount of a single object is meaningless, the overall polycount of the complete scene however, makes a difference. 

You honestly think Star Citizen is going to go to that system, let alone as a 1st year title?.

Was there ever a time CRI said that it was definitely going to the other platforms?. I thought no one here gave a damn about the "unfinished" game?.

If this has been discussed before and all that and there is a consensus on it already, never mind me =p

-------------------------------------------------------------------

Ok so earlier this evening, I was having a conversation about games sales with a friend, who is not into the topic nearly as much as myself. We were talking about what defines a console generation, with examples of things like the PS2 Slim and Xbox One X being thrown around, to try and define the boundaries. As we kept talking, we really started to focus on the PS4 Pro/Xbox One X iteration, as it's kinda doing the opposite of previous iterations. By that I mean, previous generations would iterate to simply make consoles more compact, cooler and quieter. Yes, changing to brand new chips made with a smaller fabrication process etc., but still designed to operate at the same performance specifications. Whereas, this Pro/X iteration is about adding in more performance, making the console bigger, hotter and louder, instead. Yet, there are no excluisive games for it either, so surely it has to be an iteration and not a brand new console, of course, helping us define that boundary. We then raised the point that since it's a pretty big leap into a half-generation of sorts, then the leap to the actual next generation shouldn't be too big either, right?

If we have back-to-back "iterations", with respect to performance specs, maybe consumers would be used to it enough to accept a third, then a fourth, and so on. Maybe the iterations could happen yearly, as the smartphone market seems to do. That market is quite profitable and supports some of the biggest companies in the whole world, in part due to the sheer volumes, helped by the yearly update cycles. That would make the games consoles market more profitable to hardware publishers. There would be less intense R&D costs associated with massive leaps into things like unknown architectures and multiple new technologies being added on, concurrently, each time. The companies could also get away with calling the product just "The Xbox" or "The Playstation" with yearly updates, sometimes adding one big new thing, or just refining to reduce manufacturing costs. Also, that way, they would never run the risk of what happened to Nintendo between the Wii and Wii U, where consumers didn't understand the product and how it was new. In the future, there may also be techncial reasons for this, like the eventual end to Moore's Law and limited possible increases.

Lower R&D costs, much less risk in marketing, higher annual net sales, planning for future technical risks ... Capitalism will always push change towards these objectives. 

Is it fair to predict that, maybe the next console to release in a few years ... Won't be as big of a leap as we're used to, but a move towards smaller iterations? Maybe the next time feels too soon ... but remember that the sooner it would happen, the better, for the corporations, who are in control of when this would be implemented in the first place. Maybe the PS3 to PS4 and Xbox 360 to Xbox One is the last big discreet leap (at the forefront of raw power available at the time, at least) we'll ever get from this industry.

tl;dr the next console releases won't be as massive of a leap as we have seen in previous transitions. Instead, we will eventually be served smaller and smaller iterations, to meet a yearly cycle, like the phone market.

Concerning the 10 - 12 TFLOPS. That is what I keep seeing from people on various sites. I in no way see that as PS5 levels. IMO that would be a PS4 premium/ a step above the Pro.

For SSD prices see here:

https://www.extremetech.com/computing/236260-samsung-plants-to-slash-ssd-prices-to-hard-drive-levels-by-2020

I fully expect SSD prices to be well below HDD prices by 2021, and also in massively larger capacities.

Look at some of the last ps3 and xbox 360 games compared to ps4/xbo launch titles, there was very little improvement.

I disagree. Why would you want to have just one big jump when you could just as well have regular steady improvements. People who don't want to pay money can still wait but it's overall better if we had steady new hardware releases every 2 years. Just letting people rot on their century old hardware isn't the best solution.

And with gradual improvements on hardware over the years backwards compatibility is basically guaranteed.