A recent article in Nature Medicine finds that the price of misconduct probes can surpass $500,000. Sure, that may be expensive
Researchers at the Roswell Park Cancer Institute in Buffalo, New York reached that figure by examining the monetary costs of one case at their institution, where a senior scientist was accused of fabricating images and data in a federal grant application (the case is still pending). Their results pegged the direct cost of the investigation thus far at around $525,000
Cost of poor research: I have never found any figure although costs may be even in the range of 10- or 100 fold either by provoking unnecessary studies or by delaying necessary research.