I would expect to see such failed replications even if LK99 is the real deal. Getting the exact conditions correct for any sort of novel lab science is non-triv

I would expect to see such failed replications even if LK99 is the real deal. Getting the exact conditions correct for any sort of novel lab science is non-trivial, and often times the crucial variables might not even be measured in initial publications. People use this fact to try to bash things like cancer research, but it is unfortunately just a fundamental problem of science. Scientific publications are not kernels of distilled truth, they are work-in-progress commits. If we waited to publish until every last detail is known, science discovery would slow to a snails pace and we would miss a ton of discoveries.