Would this not incentivise the forming of groups that replicate each others work. If you're already committing wilful fraud on your own papers, why wouldn't you commit a bit more for another researcher willing to do the same for you? With >2 parties, it won't be immediately obvious that this trading has occured.