Do reads at high concurrency get a boost if I where to raise this value?

  <ConcurrentReads>8</ConcurrentReads>
  <ConcurrentWrites>32</ConcurrentWrites>

Any benchmarks/reports showing a good sweet spot per CPU core?