I think I see what MS is trying to do here. My guess is that they want something that looks more like the mobile phone model for consoles. Which is to say, rather than the "hard" generational breaks you get with the traditional console cycle, where every 5-8 years a new console comes along and renders the old one obsolete, they instead want new hardware every 2 years or so (at a guess), which emphasises evolution rather than revolution.
What I also suspect is that they're planning a kind of limited back/forward compatibility system for games. They've repeatedly said that Scorpio will not get exclusives. A lot of people are suspicious of this, but I actually believe what they've said. That said, I still think they're being disingenuous. Their next step will likely be another console iteration maybe 2 years after Scorpio (2019), let's call it Sagittarius, whose titles will be playable on Scorpio hardware, albeit with lower performance, but not on the current XB1. The eventual successor to Sagittarius (2021) will share compatibility with that console, but not with Scorpio - and so on. So Scorpio will technically never have exclusives.
That said, this is still a risky proposition. By and large, console gamers like the fairly long console cycle. They're usually on a tighter budget than PC gamers and being able to get away with very infrequent hardware changes is a plus.
Moreover, what this plan (if it is indeed their plan) would do is eliminate the mid/late part of the traditional console cycle. That's not necessarily a good thing. For gamers, the early part of the cycle is usually a pretty dire time. Early adopters tend to get a mixture of thin technological showcases and sloppy, hurried ports of games originally developed for the previous generation. There are very, very few classic console games that were early-cycle releases, from the mid-90s onwards. In the mid/late cycle, developers are comfortable with the hardware and the focus shifts more onto the actual games.
The mid/late cycle is also traditionally a good time for the console manufacturer. Launch windows are awful. They're risky and they need a lot of upfront investment (in hardware development, games development, support for third parties and marketing) that can be hard to recoup quickly. By contrast, in the mid/late cycle, the real cash cow, which is to say the third-party licensing fees (which are, I cannot emphasise enough, where the real money is in the industry) are flowing in nicely. Admittedly, in the 360/PS3 generation, the late-cycle was allowed to go on too long and gamers lost interest, but that was more down to tactics than industry structure.
So in some respects, this looks a bit of a self-destructive strategy. However, I think the industry has painted itself into a corner in this generation. For the first time I can remember, the real battleground between the main rivals was not their exclusive games franchises, but on multiplatform performance. With modern development costs, platform manufacturers can no longer afford to fund the same number or quality of outright exclusives. Instead, the PR battle was fought on technical specs; Sony annihilated MS when the PS4 and XB1 launched because the PS4 had some nominal performance advantages that were hard to even perceive for most gamers, but which made great marketing.
So the industry has locked itself into a battle of technical one-upsmanship. Worse, it's done so at a time when PC gaming is seriously resurgent. Trying to get into a tech-specs battle with the PC gaming scene is an unwinnable fight. So now, if Sony and MS don't want to lose a fight on the ground they themselves have chosen, they need to keep iterating the hardware to remain competitive.