I absolutely loathe this idea. It's the same line of thinking that led to the unified AEDU system in 4E, and I'm a bit surprised Paizo would get drawn into it. Among the folks who abandoned 4E for Pathfinder, AEDU was high on the list of things that they hated like poison.
'Hated' is the key word. The rejection of 4e by h4ters was not an entirely rational reaction. There were a lot of factors driving it, from the perceived 'early' rev-roll, to the behavior of WotC, to rejection of change, to feeling betrayed by the invalidation of hard-won system mastery, and many more rationalizations, excuses, talking points, even the odd real as well as imagined slight here or there. It was a perfect storm, of sorts.
Leveraging an existing system that is already imbalanced and already mastered by optimizers, is thus a non-issue, even if it does, in fact, constitute one of the proxy issues that h4ters rallied against in the edition war. Pathfinder is imbalanced and lavishly rewards system mastery, already, so it can afford to introduce a form of psychic magic efficiently by leveraging existing systems. Besides, Pathfinder is extremely complex and bloated, already, so they have to be thinking about keeping it playable.
5e is also imbalanced, but it's rewards for system mastery haven't risen to the level of 3.5/Pathfinder, for want of 15 years of bloated material. So it's debateable whether it'd better-serve the h4ter sub-culture of D&D fans to bloat 5e out , further imbalancing it and adding the kinds of needless complexities, unintended synergies and broken combos that reward system mastery, or to leave it at it's current level of imbalance and exploitability to hold the line on a professed goal of relative 'rules lite' simplicity, which seems to be appealing to the sub-culture of Classic D&D fans (even though most of classic D&D - all of it but, perhaps, B/X - was actually quite complicated).
People on forums talk as if there is a bright dividing line between "crunch" and "fluff,"
There /can/ be, it's all a matter of how the game is designed and presented. In 4e, there was such a dividing line. Fluff was in italics in one part of a text block, keywords and other crunch in regular or bold type in other parts. In classic D&D, the two were mixed so freely and thoroughly it was hard to say either really existed - 'rules' were as often phrased entirely in the manner of vague, evocative 'fluff' with no consistent interpretation possible, 'fluff' could incidentally take on the character of a rule, mechanics could be seen as defining the setting, and so forth. In 5e, rules are written more like they were in classic D&D, but spells, at least, do have a descriptive section preceding a more rules-focused one - a cosmetic 'compromise' that is functionally no different from the classic approach.
If two classes use the same rules, they will feel "samey" even if they are described as wildly different. If psionics is to feel different from magic, then it needs to use different rules.
Not remotely true, but an understandable misconception. It is easy to differentiate two game elements by giving them radically different mechanics - at the price of increasing complexity. It takes a little more thought and subtlety to differentiate mechanically similar elements - but doing so limits bloat, broken combos, and the like, and makes future design easier. So it's a balancing act in how the lead developer uses the available design resources.
The nice thing about D&D is that if you don't want to learn the psionics rules, you don't have to! Just stick with wizards and sorcerers. The DM should have a vague sense of how psionics works, but the only person who needs a really firm grasp of the psi rules is the person who chooses to play a psion.
That trick worked in 4e, when basic mechanics were consistent, and it was easy to evaluate a power or class ability in the moment, with each player really only needing to be really familiar with his own character and the DM able to focus on the encounters. In all other editions, and particularly in intentionally-DM-empowering 5e, it is critically important for the DM to know everything the PCs can do, at least as well as the players know it. Heck, in 1e, EGG advised the DM know all the rules /better/ than his players, and if he didn't, that he'd 'lose control' of his campaign.
That's a challenge, but it's one most long-time DMs are well able to handle.
All that said, I think it's a wash. Mr. Mearls is relatively free to handle psionics as seems best or most expedient. That 5e doesn't seem poised for the usual flood of splatbooks, and WotC farms out so much of the little that is published for it, though, argues for limited design resources, which makes a UA sub-class seem more likely than a complicated separate/mulitiple- sub-system splatbook. FWIW.
tl;dr - Neither Pathfinder nor 5e face an edition war, so either or both could afford to re-cycle existing sub-systems for psionics, even at the risk of mechanical 'sameyness.'