That’s a tall order! Even only the first two points are hard to meet at the same time, as they effectively contradict each other. Virtualy anything that devs or the industry needs already exists somewhere. Especially if it is easy to implement (4th point). And if a solution exists in other languages but is not optimized, then maybe there is not much need for an optimized version (point 3). So all of the points are hard to meet at the same time.
A good starting point could be to “scratch your own itch”, that is, start with a problem you face personally and write a solution for that.
Or how about contributing to already existing projects as a start? New problems that justify a new OS project might become visible from such interactions.