Anything that is new that is unbalanced could have been designed balanced.
If something is unbalanced, that's the designer's fault for not putting things in the parameters accepted as balanced.
Only in the same sense that any new software
could have been designed and implemented without bugs, and it's the programmers' fault that it isn't bug-free. Superhuman game rule designers able to see all the implications of their designs in advance are no more readily available than superhuman programmers able to see all implications of their code in advance.
(Seriously, consider all the time and effort that went into the design and playtesting of 3rd edition, and then look at the mess that its magic system was proved to be the moment people stopped using the play style that had formed under 2nd edition and exploited the rules to the fullest. Cook, Tweet, and Williams were not incompetents, and neither were all the many, many playtesters.)
D&D spell magic as eventually embodied in 5th edition is the product of many cycles of design, testing, patching, and adaptation undertaken over four decades. Unless a not-based-on-spells psionic system is explicitly made narrow and limited (and thus disappoints psionics fans), it
will be more broken than 5e spell magic.
But that doesn't matter, because if the new psionics were, by a miracle,
only as complicated and broken as 5e spell magic, it would be rejected by most tables. People have already learned spell magic, and already are inured to its broken bits. Most groups as a collective will not want to learn an additional system as complicated as spell magic, and when the others try it out, every time they stumble across a broken bit it will irritate
because it is a new issue.
Thus, from the perspective of a publisher, a not-based-on-leveled-spells, non-trivial psionics system that works as well as spell magic would take a massive investment of resources into developing and testing something that would be widely rejected. As an economic proposition, that's inherently losing.