Non-phantom Dark Energy and a Massless Lightest Neutrino: A Joint Pre-registration

This paper combines three judgements into one joint hypothesis and fixes the decision rules before the data arrive. Hypothesis 1: dark energy is non-phantom — as a single component its equation of state satisfies w ≥ −1 at all times, with equality for vacuum energy. Hypothesis 2: the universe is CPT-symmetric about the Big Bang, one consequence of which is that the lightest of the three neutrinos is massless and neutrinos are Majorana (Boyle, Finn & Turok 2018). Auxiliary assumption: the mass ordering is normal (the current preference of global oscillation fits). From Hypothesis 2 and normal ordering, the neutrino sector has no free parameter apart from a Majorana phase: Σm_ν = 0.0588 ± 0.0002 eV, and the effective Majorana mass for neutrinoless double-beta decay lies in m_ββ = 1.50–3.72 meV. Following the reading of earlier work, the tensor-to-scalar ratio is r < 10⁻³. None of these numbers is new on its own. What this paper adds is a directional judgement that follows from the combination. Under ΛCDM, DESI DR2 together with the cosmic microwave background (CMB) already pushes the neutrino mass sum close to, or below, the oscillation floor (about 3σ tension for the effective-mass posterior). What loosens the bound appreciably is evolving dark energy that crosses −1; non-phantom evolution gives bounds comparable to ΛCDM. Hypothesis 1 closes the first route. We therefore expect the tension to be resolved by the measured mass sum moving upward and settling near 0.0588 eV, and the evidence for crossing −1 to weaken rather than strengthen. In the reading adopted here, "non-phantom" and "no high-scale inflation at the beginning" are the two projections of a single condition — that the end of the universe and the beginning of the next stage match across their junction — and w ≥ −1 guarantees that the area of the cosmic horizon (its entropy, the upper bound on information capacity) never decreases, and is required for this once dark energy dominates. For each prediction we give the object, direction and value, deciding data and timing, pass line and fail line, and we fix in advance how adverse outcomes will be read and which rescues are forbidden. The earliest deciding data are the cosmology results from the DESI five-year sample (expected 2027; DESI Collaboration 2026); Euclid’s first data release is published in stages from November 2026 (ESA), with cosmology analyses to follow. Sealed text SHA-256 (Sections 5 and 7): c9fd06b58abe3ee1b4d4a5b7c9d34264d48fd31870775abe5f564a68b8e100bc