What we don't know, published on purpose
“Never a homemade number” means something specific: when the evidence has a hole in it, we say so instead of quietly smoothing over it. Below are every documented gap behind the CoverScore registry — 15 total, 14 still open.
aging · 7 gaps
Age-65 lifetime Alzheimer's risk by sex from the 2026 F&F edition (p.32) not extracted; only the age-45 anchors (1 in 5 women / 1 in 10 men) are verified.
SSA Supplement Table 6.B5 sex-specific claiming shares could not be reliably extracted (two fetch runs conflicted). Totals verified; do not use by-sex splits without a direct table read.
No source publishes 10-year death probabilities at ages 50/60/70/80. Only 1-year q(x) and l(x) exist; a decade figure requires a labeled derivation.
FBI IC3 publishes no age band above 60+; BJS identity theft none above 65+; FTC's Protecting Older Consumers aggregates 60+ except 80+ callouts.
LTC insurance ownership rates by decade band (50-59/60-69/70-79) not published; only 50+ and 65+ aggregates.
No federal per-capita nursing-home residence rates by age since 2004 (NNHS discontinued; NPALS publishes resident distribution only).
Historical SSDI incidence per 1,000 by 5-year band from Actuarial Study 125 Table 4 (p.23).
Closure note: 2022 Trustees Report ultimate incidence rates obtained (life.ssdi.incidence); Study 125's historical Table 4 remains unextractable (PDF truncation + proxy block).
contents · 1 gap
No study anywhere compares settlement amount or speed for policyholders WITH vs WITHOUT a pre-loss inventory. Confirmed absent across UP surveys, state DOIs, NAIC, academic literature. Do not invent; use the burden framing (cont.itemization.burden).
data privacy · 2 gaps
No causal effectiveness study of credit freezes on new-account fraud. Mechanism argument only; nearest evidence is the Philly Fed extended-fraud-alert paper (2025).
No P(financial misuse | breach exposure) by data type (SSN vs credentials vs medical). GAO's last federal attempt (2007) found it unmeasurable.
health · 1 gap
No published NNS for cervical, hypertension, diabetes, or lipid screening; no absolute ARR for FIT/gFOBT/sigmoidoscopy in current USPSTF materials.
life financial · 1 gap
SSA publishes no probability of disability/death from an attained age (50s/60s) to NRA — AN 2025.6 tables are cumulative from age 20 (verified against table headers).
property · 2 gaps
No agency publishes annualized structure-loss probability by CAL FIRE FHSZ tier; First Street 30-yr vendor tiers are the nearest substitute.
No official 30-year flood probability for FEMA zones X / X-shaded (the 1-in-4 figure covers SFHA only).
vehicles · 1 gap
No public at-fault collision probability table by age band; only per-mile ratios and crash shares. Claim frequency by annual mileage is telematics-proprietary.
This page updates as the evidence base grows. If a gap above closes, the registry — not marketing copy — is what changes first.
Read the full commitment →