Replication is kind of fuzzy though. A lot of things are replicated, just indirectly, building on suppositions supported by but not directly reproducing the stu

Replication is kind of fuzzy though. A lot of things are replicated, just indirectly, building on suppositions supported by but not directly reproducing the study. The problem here is perhaps one of reporting negative results. It'd be very easy to rationalize that a secondary study didn't work because, hey it's not direct reproduction and this use case is invalid or the equipment available wasn't the same and etc...). And between the rationalizations and the lack of incentive to submit and publish negative results it's easy to extrapolate that an indirect challenge may take a really long time to be put forward. Something kin to a bug report section might be in order. A "Works on my machine." regime isn't acceptable in the empirical domain. This would help prune dead ends in both primary and secondary+ research, one would hope, without necessitating direct redundancy. Of course this may run against the grain, fears of getting scooped or IP challenges and etc...