If the "failed replication" was a single study, as in many cases listed here, there is still an open question as to whether the 1) replication study was underpowered (the ones I looked at had pretty small n's), or 2) the re-implementation of the original study was flawed. So I'm not so sure we can quickly label the original studies as "debunked", no more than we can express a high level of confidence in the original studies.

(This isn't a comment on any of the individual studies listed.)