Thanks -- very thoughtful and helpful. Other than spending some time figuring out the optimal indices, what you propose sounds a lot like what we're about to register. I think that my post did not make clear the big red line you're also alluding to: we have no intention of not doing anymore analysis that either departs from pre-analysis plan or digging into insignificant overarching indices that have plausible heterogeneity of impacts in sub-groups. It's just that we want to set them apart to ourselves and the readers.
We actually think that there will also be a fair amount of methodological secondary analysis -- e.g that have to do with extensive vs. intensive margin effects, in interpreting the findings, etc. These will be part of the paper (or 'science' as you call it) and would distinguish the work from a rote reporting of a pre-analysis plan -- which is useful but in a limited and sometimes uninteresting way...
Thanks again for taking the time to comment.