The "market fragmentation" and "poorly timed" arguments are missing the point (though "PS4 Pro" isn't a very smart forward-looking name). This isn't really the console market anymore. Sony and Microsoft have finally consolidated their platforms enough that they can finally just use commodity PC hardware, but don't be fooled - the big players are under pressure from larger market forces. In a relatively short period consoles went from offering superior and different capabilities than PCs - at very competitive prices - to now being rapidly depreciated hardware without much to offer except exclusives and perhaps some slight price and convenience advantages. This move simply allows console gamers to take advantage of many things PC gamers have taken for granted for years, while still keeping cost and convenience advantages.
PC gamers have known for a long time about price segmentation, and it's not a bad thing. It doesn't necessarily lead to market fragmentation. It's no big deal that new graphics card architectures launch a full lineup, not just one option; buy something extravagant now, or wait a while to get the same or better performance cheaper. Around the time of the PS3 launch, somebody at Sony said they intended to keep the "format" alive forever. PS2 compatibility was shortly downgraded and eventually dropped for the same reason the SNES didn't include NES compatibility - it wasn't economically feasible. Emulating the PS2 on the PS4 was an "extra," so Sony charge for it (ditto for Microsoft). But moving to universal hardware and forcing better software practices actually allows the format to survive as long as X86-64 gaming R&D keeps going - going forward both Sony and Microsoft will be keen to jump on any moves by their competitor to break faith with this new arrangement; forcing a compatibility break will be a hard sell. That's also good for developers whose sales may have a longer tail. (In fact, Microsoft looks to have been fairly close to getting here except that they moved to PowerPC for the 360 shortly before Apple moved away from it. It is funny to consider that this invites a return to the simple Atari-style numbering schemes, when newer systems simply had arbitrarily chosen names to differentiate them from what came before; Microsoft's more neutral "Scorpio" sounds better than "Neo" in this regard, at least).
A couple other things are obvious: Improvements in performance per watt / per die size area / per clock have all slowed down lately. I don't know if this is a permanent reality going forward or whether moving to new materials like Gallium compounds will allow faster clock speeds, but right now it appears to me that many performance increases will be slow in coming (putting aside the inefficiency of the current AMD X86-64 architectures prior to Zen). The simple result of this is that you probably don't need to fear annual console updates, though I'd question whether that's actually something to be fearful of.
The apparent failure of Kinect and the ho-hum response to VR thus far also appear to be relieving product planners and gamers of the pressure to try and leverage "new gaming paradigms" to sell systems, when those sales were really more driven by the ecosystems (community and existing games).
And finally, depending on what the new backwards compatibility standards - like PS4's "base" mode - hold, we might see technical improvements in games. For certain PS4 Neo titles must render at 1080p, minimum, instead of 900p upscaled or some rubbish. I don't see any information about framerate targets, but I couldn't overemphasize how important something like this would be towards making consoles first-class gaming systems again. Of course,
Eurogamer questions whether some of the requirements couldn't actually cause competitive handicapping of one system compared to another, so the details matter, at least if you're interested in multiplayer or score competitions.
. So perhaps a Daytona esque title would be able to push the Pro to its limits.