The reproductions showed:
1. The ingredients commercially available have iron impurities. The oven baking process causes the iron to separate out, often towards the edges of the material chunks. Iron is ferromagnetic.
2. The LK99 material itself is strongly diamagnetic, which is interesting in and of itself, but obviously not world-changing.
The result of the combination of the above two is that the iron impurity is attracted to magnets, but the rest of the LK99 material is repelled. This makes samples "stand up", but not hover like a true superconductor would. All of the third-party reproductions showed this effect, none showed samples truly floating.
Similarly, the conductivity tests were zeroed badly. True superconductors have a step-change in conductivity, whereas the graphs measured by the korean team showed a rapid change followed by a slower gradual decrease. The lowest resistance they measured was still very high, comparable to metals, not superconductors.
This is a common issue with fringe research into superconductors: the "signal" they're looking for is a step-change in conductivity because it is easy to measure and rare in most normal materials. However, sudden conductivity changes can occur for all sorts of reasons. It's necessary, but not sufficient for superconductivity to be confirmed.