We're not killing them off, they're just failing to adapt and advance with the rest of MAME because as Tafoid says, the ones you're praising are all based on very old code and haven't been properly maintained / updated over the years.
As with any support software it is important that it is responsive to the needs of the project it is meant to support, it's a case of adapt or die.
When the support projects simply bury their head in the sand and ignore progress, or attempt to hack the progress out in order to keep their own software running it's inevitable that one day they will fall apart. Unfortunately that is what has been happening with the frontends you talk about; instead of embracing MAME some of them have been more or less actively working against it, and after a while that becomes unmaintainable.
Projects like QMC2 on the other hand have embraced change, have seen where the project is going, have even been proactive in their support, shown great foresight, anticipated what will happen next and planned well into the future with their design. You might not like them as much, but the reason they're recommended, and still fully functional is because more thought, planning and ability has gone into their creation and recent maintenance. Obviously they have an advantage here in being relatively modern, as opposed to being written at a time when nobody really knew just how good MAME would get and what it would be capable of, but nevertheless it remains a true statement.
All you're seeing is legacy code and poor choices catching up with the frontends you want to use.
It's the same for any software fwiw, code always ultimately hits limits and points at which it needs replacing / rewriting rather than just hacking away at. It has happened in the rest of MAME many times, if you look at the project as a whole it is very different now than it was a few years ago, and even with that in mind there are complex and challenging problems we need to deal with in order to advance, things that at the time we didn't consider to be important but are now actually big issues.