Benchmark: running BigRE, Index, OptRE, m// loop, each for at least 10 CPU seconds...
     BigRE: 10 wallclock secs (10.59 usr +  0.00 sys = 10.59 CPU) @ 77.59/s (n=822)
     Index: 11 wallclock secs (10.50 usr +  0.00 sys = 10.50 CPU) @ 1427.58/s (n=14991)
  m// loop: 11 wallclock secs (10.47 usr +  0.00 sys = 10.47 CPU) @ 686.50/s (n=7187)