I don't see the need for publishing the constituent studies before publishing the "meta-analysis". The individual studies aren't contributing unique theoretical arguments; they are additive toward a single argument that the original results can't be replicated. It's fairly common for a single paper to incorporate multiple experiments (the paper is closer to this style than a traditional meta-analysis) which are all addressed by the same reviewers.
Good points and your comment and the other sibling comment definitely make the same valid argument. I got hung up on the phrase meta-analysis, which does seem to be misused, or at least not used in the typical sense (which as was noted in the post was brought up by a reviewer as a problem). It's not a meta-analysis of existing research, it's purely new research to present multiple replications attempts, all of which failed.
