Yes, this idea has been explored. See the paper FlexSC: Flexible System Call Scheduling with Exception-Less System Calls (Soares & Stumm 2010), https://www.usenix.org/legacy/events/osdi10/tech/full_papers.... In addition to having async system calls, they dedicate an entire core for processing system calls to the kernel, in order to minimize CPU cache thrashing. They've implemented this idea on Linux and sped up Apache by 116% (so 216% the original performance), MySQL by 140%. Unfortunately it hasn't caught on in production anywhere.
Nitpick: they improved MySQL performance by 40% (so 140% of original performance), which is still very impressive.
It's also amazing that they were able to make it work transparently for applications. The only way it would catch on is if it were upstreamed though I guess.
