I'd assume that depends on your measure of worth, I'd say. Many operating systems had little academic significance when it was most academically or commercially fruitful to invest time in. Microkernel and dependency specific operating systems would be interesting. Or hardware based capability based operating systems.