A typical paper may cite dozens or hundreds of other papers. This does not sound feasible. It honestly seems like it would worsen the existing problem and force

A typical paper may cite dozens or hundreds of other papers. This does not sound feasible. It honestly seems like it would worsen the existing problem and force scientists to focus even more on their own original research in isolation from others, to avoid the expense of running myriad replication experiments that they likely don't have the funding and personnel to do.