Research gap intelligence
What we don't have, and what it would change
Every entry below names a specific missing evidence object, why it matters, what
acquiring it could change, and — the part that keeps this from being a to-do list — what it could
not establish even if found.
Blocked questions
Questions the corpus currently cannot answer, and the gap that blocks each.
Proposed research targets
Priority is derived from stated dimensions. Feasibility is deliberately
excluded from the priority rule — difficult access does not lower research importance.
Admission to the corpus
The only path into the canon. It has seven links and four separate authorities —
a research need, a worker, a verifier, and a human — and it is deliberately boring: a human
authorizes one exact object, and a mechanical writer copies fields without deciding anything.
ADMISSIBLE ≠ AUTHORIZED ≠ ADMITTED.
Verdict lifecycle
A verification decision is immutable, and a package may carry several of them.
Which one has authority is a separate, explicit, human-authorized act — there is no
rule here that says the newest wins. Where several verdicts remain unsuperseded, this page
says AMBIGUOUS rather than picking one.
Controlled research run
One human-approved target, one bounded run, one proposal.
There is no arrow from a proposal to the corpus.
Verification
An independent verifier examines the package. It cannot repair the proposal,
and it cannot admit anything to the corpus. Four different things are reported separately —
conflating them is what would make this pipeline worthless.
Acquisition routes already tried
Failed routes are research intelligence. Nothing here is discarded.