A prediction market can help setting priorities on replicating research. When many predict a scientific result will be falsified, this is a clear red flag. With high stakes, there is also an incentive for some market participants to do the work themselves. Clearly, prediction markets are then not a substitute for research. The very existence of a warning system like this can make new research better to begin with too. Perfect? No. Perfect now? Neither.
Derived from own comment on Prediction Markets for Science: What Problem Do They Solve?