This is a common misunderstanding. The main mechanism for self-correction in science isn't about replicating or refuting findings. It's about which science survives to become the basis of other science. If a finding does not bear out, it simply will not become the basis for subsequent science. It's similar to natural selection.

The so-called "replication crisis" is overblown.

If LK-99 fails to be a useful material, it's interesting life will be over very shortly. If it is useful (and even if it's not the promised room-temperature superconductor, but useful in some other aspect), it will continue on in science, in technology - everywhere it might be useful for further investigation.

...the replication crisis is not in science