In code the reason we don't do wholesale rewrites of complicated systems very often is because while there are bugs in the current system they're known quantities; going back to the drawing board and starting from scratch results in a newer, better, but unknown system with unknown bugs. That might be worse. It's a massive risk.

This is true for Apple's new keyboard too. We can't automatically assume it's better than the existing one, or that it has no issues. It's an unknown. For all we know a new keyboard might be even worse than the current one. Hopefully Apple will have learned from their mistakes but whether or not they've corrected all the problems or introduced new problems is something we can't know.

It's good that Apple appear to be listening and are trying to fix the problem (if that's the reason for the new keyboard) but I won't be rushing out to buy one for a while.