# n148 — the speed lock, first assault: INCONCLUSIVE (recorded)
# (2026-07-11. Not a pass, not an execution - a measurement that came
#  back too weak to decide, with a diagnosis that deepens the story.)

## The attempt
Global fit of the Euclidean shear bubble (m = 128, 81 grid points):
analytic polynomial (the lattice/bare side of n145's cancellation)
+ C(Om^2 + c_g^2 q^2)^{3/2} (the would-be universal term), c_g free.
RESULT: the non-analytic term is UNDETECTABLE - fit improvement only
x1.3 over pure polynomial, c_g runs to the scan edge (3.4), rms
nearly flat in c_g. S1 FAIL as declared; S2 unmeasurable.

## The diagnosis (the protection and the lock fight over one object)
The universal non-analytic term's strength is the imprint of the
on-shell pair continuum - and n140 measured that the shear channel's
edge weight VANISHES as (omega - cq)^{2.7}. The same zero that
protects the collective graviton (no decay width on the cone, n147
T3) SUPPRESSES the non-analyticity the speed test was hunting: the
shear channel is nearly analytic in the IR. Consequence: the naive
(3/2)-power model was wrong, and worse, the mode's dispersion is
dominated by ANALYTIC terms whose Om^2/q^2 ratio is not obviously
Lorentz-protected. The speed lock is neither closed nor executed:
it is REOPENED at a deeper level.

## Next (named): the correct extraction
Derive the true non-analytic power from the edge exponent
(rho ~ (E - cq)^{2.7} integrates to a softer branch cut), redo the
fit with that power at larger m - or find the symmetry argument
that ties the analytic coefficients themselves (the cancellation
structure may enforce the ratio; n145's two negative dressings and
the bare term all renormalize the SAME operators). The campaign's
live risk stays live, now with its precise shape known.