It's not really the microkernel per se that delayed GNU/Hurd so much. From the beginning the Hurd project had a more ambitious interpretation of the microkernel

It's not really the microkernel per se that delayed GNU/Hurd so much. From the beginning the Hurd project had a more ambitious interpretation of the microkernel paradigm than other microkernel operating systems (e.g. Minix), and this shifted the project from engineering to research, leading to never-ending architecture changes and even wholesale microkernel replacements: https://www.gnu.org/software/hurd/hurd/ng.html https://www.gnu.org/software/hurd/hurd/ng.html