All of this can be simply fixed. Either by journals (only publish replicated work, other papers only publish the abstract) - not likely - or create an index of

All of this can be simply fixed. Either by journals (only publish replicated work, other papers only publish the abstract) - not likely - or create an index of tainted papers. A paper that is tainted because can't be replicated by others, also taints all papers that quote it. Index can be searched by author name. People would be much more careful what they quote. Basically without replication you do not get quoted. Without being quoted you do not get scientist karma.