the problem is new sw also didn't fix the last 10%,
solution? add the last 10% to the existing solution (possible with open source) without reinventing the wheel completely, just to be able to stick a flag in the mud
"You wanted a banana but what you got was a gorilla holding the banana and the entire jungle."
Sometimes, it is better to have something that solves the 10% of the problem you are facing than something that solves 9 other problems you don't have but forces a cognitive overload on you because of the more general abstraction for dealing with those problems. Usually, you cannot solve all problems for all people and you need to make trade offs.
Reinventing the wheel is really a bad analogy for this. A car wheel is not a substitute for a bicycle wheel, a bicycle wheel isn't much good on a wheelbarrow, a wheel barrow doesn't help out much on a skateboard, on and on and on. The act of inventing the wheel defined its parameters and necessary characteristics so that (mechanical) engineers could have them manufactured to their specifications as needed for the thing they were designing. If I design a new wheel it isn't invention.
The similar "fit for purpose" notion in software is not widespread. Many of us somehow feel that implementation is more invention than manufacturing or design. The goodness of fit criteria are not often discussed or are difficult to define or measure. The things that sqlite can do that postgresql cannot and vice versa are important. They each leave some problems unsolved so that they can solve their chosen problems.
solution? add the last 10% to the existing solution (possible with open source) without reinventing the wheel completely, just to be able to stick a flag in the mud