It is hard work, but relatively easy work proportional to the possible rewards, and has very few downsides:
1. An early preprint provides instant press coverage for the researchers, their lab and the entire university. Such opportunities are rare: I worked in academia full-time for decades, and never had an opportunity to do some fixed amount of well-understood work that guarantees an appearance in the national press. Such an appearance is very useful in inter-departmental politics, not to mention when dealing with government organizations staffed by non-researchers. E.g. if "biggest newspaper in country X" reported about your research, that by itself ensures that you won't have to apply for "shitty academic visitor visa to the UK", but can credibly apply for a "global talent, go straight to indefinite leave to remain" one instead. And if the effect turns out to be real, and you managed to reproduce the effect while lots of others tried and failed, that's going to look really good on the grant application where you have to explain why your lab is the best place to spend money earmarked for superconductor research. This is unlike other replication papers, which are not usually the "gets reported on national media kind, or even "many others tried and failed" kind, but usually the "nobody else cared" type.
2. The article will definitely garner some citations. Even if LK-99 does not pan out, there will be many many survey articles written about what's unfolding right now, and they will all definitely cite the earliest replications. Moreover, it doesn't matter if your first preprint has shit writing: you'll still be cited if the final version gets accepted 4 months from now, after many rewrites. This is unlike most random "reproduction" papers, which are unlikely to get published, much less cited. So reproducing LK-99 is in fact a good way to increase h-index.
3. As long as you don't do anything fraudulent, there's very little career risk involved with being fast-and-loose in a preprint. If LK-99 pans out you're definitely in the clear. If it doesn't, well, as every second comment here mentions, materials science is difficult, honest mistakes are easy to make, even reproducing results is incredibly finicky and hard. It's not like your lab was the only one making anomalous observations. Chances are nobody's going to care if you were involved in a false positive replication. In fact, probably nobody's going to read and scrutinize your whole publication list when hiring or promoting you, but they'll definitely care about your citation metrics. This is unlike most other reproduction papers, where you wouldn't get many citations even if you managed to publish, because nobody cared.
All in all, dropping everything else and working on this right now is a good strategy, even for investigators in prestigious labs that would otherwise not bother with reproduction papers.