This happens so frequently to me. I'll Spend an hour running benchmarks and then post a PR. Someone chimes in and says "method x is faster" with nothing to back it up. I have to benchmark again, and then my way is still faster.
In any case you can just say: "OK, until someone comes up with evidence to the contrary, I think we should go with my version. If someone shows up with a better version later, I have no objection to merging that."