From the threads I read about the Psi die in various places, and listening to this, I think that they really shot themselves in the foot with the die in two ways:
1) The "out-of-power" mechanic where the die could stop working. People absolutely hated this.
2) The fact that the die went up and down in power at all. This vexed the heck out of people, though nowhere near as much as the die stopping working.
The math showed that these reactions weren't entirely rational (just like a lot of the dislike of the Mystic relied on preposterous situations, mathematically), but they were nonetheless common, and no amount of people doing the math could convince people otherwise.
Crawford discusses how some people don't like unique mechanics, but I feel pretty confident in saying that if they Psi die had been a fixed number based on your level (that could perhaps be boosted/lowered, but only by "voluntary" effects, not random ones), it would have done two things:
A) It would not have been regarded as a "unique mechanic". Loads of classes have some sort of die mechanic, but none of them have it randomized. That's where this crosses into the line of new mechanics.
B) It wouldn't have annoyed so many people with the declining/increasing aspects who otherwise didn't mind it as a mechnanic.
Had they just done that, yes, it would have been different from what they got, but the combination of it not being regarded as a entirely new mechanic by most people, and not being as potentially vexing (in that it clearly made some people anxious, and others just didn't like the idea), then they'd have got a thumbs up on this. Though they'd still have not had enough mucus for the Sorcerer! So that would need fixing.
On the upside, Crawford seems very positive about including Psionics, like he really repeatedly stressed how it's always in D&D, over and over. So I expect we'll see yet another take on these guys pretty soon.
If they do insist on using existing spells and so on (which I don't think is a good idea, because it annoys too large an intersection of groups), I think the main thing is to make it so the system is not slot-based (they already a spell point system in the DMG for goodness sake!), and make it so that by default, you cast w/o any material components at all, and and probably choose whether you use V or S components (pick one, other vanishes), and maybe even can suppress whichever of those you have for a small psi point cost.
Unfortunately this is I think of one of the D&D design team's foibles. They consistently do a few weird things - I've mentioned that they wildly overvalue natural weapons and natural armour, like to a bananas degree. It's inexplicable. In almost every case they make them a terrible idea, and objectively vastly worse than say, having Elven Weapon Proficiencies or the Mountain Dwarf deal with okay weapons and medium armour, yet they value them as if they were the same as, or better than that. I have no idea what sort of games they're playing, but in thirty years of D&D, the only "natural weapons" using characters I've seen have have been either gimmick characters, where it wasn't mechanically advantageous, just cool, or weird exploits (like where some race got to retain a claw/claw/bite routine or whatever), and the latter just isn't a thing in 5E.
And I think likewise they overvalue trying to make people using full V/S/M components, even though, realistically, 95% of the time, it doesn't even matter, and the other 5% it's either costly components (just adjust those spells or take them off the list) or casting from steatlh - and psionics should be good at being cast from stealth. That's part of it's "thing". So I think they'll mess it up if they try to go the "spell" route.
