It doesn't make sense to run benchmarks in our sanitizer variants (A,UBSAN Enterprise RHEL 8 ARM64 DEBUG (Benchmarks)). Sanitizers completely change the performance characteristics of the benchmarks and the runtime isn't meaningful. To save cost and patch runtimes we should remove them from the variant.

Our correctness tests in jstests are a strict superset of what our benchmarks run, so we won't lose any sanitizer coverage.