============================================================================
PLATEAU CROSS-CHECK + FLOAT THRESHOLD + VERDICT
============================================================================
Simulation vs closed-form psi* = c * annuity(g, dur):
  c=0.39% dur=40y: sim 10.9% vs closed 10.9%  (match)
  c=0.39% dur=30y: sim 8.9% vs closed 8.9%  (match)
  c=0.50% dur=40y: sim 14.0% vs closed 14.0%  (match)
  c=0.30% dur=25y: sim 6.0% vs closed 6.0%  (match)

Active float left at the plateau (price-discovery margin):
  central 10.9%: active float = 89%  (discovery concerns bite ~40-50%+ passive)
  high corner ~21%: active float = 79%  (discovery concerns bite ~40-50%+ passive)

============================================================================
HONEST VERDICT
============================================================================

 This one largely VINDICATES Paper 8, with one nuance.

 1. The plateau is REAL (Proposition 4 holds). Ownership converges and never runs
    toward 1 in any swept case -- cohort decumulation structurally bounds the share.
    The 'permanent buyer eventually owns everything' objection is genuinely defeated.

 2. The simulation reproduces the paper's psi* = c*annuity(g,dur) identity exactly,
    and the central plateau (verified 0.39% flow, ~40y duration, 2% growth) is 10.9%
    -- squarely in the paper's stated 0.09-0.11 bracket. The number is sound.

 3. NUANCE (the honest addition): the plateau LEVEL is duration- and flow-sensitive,
    ranging ~6% (short duration / low flow) to ~21% (long duration / high floor-
    weighting). The paper's ~10% is the central case of that range, not a fixed
    point; at long holding durations it could reach the high-teens to ~20%. Still a
    clear minority -- active float stays ~79-94% across the whole range, well above
    the ~40-50% passive-ownership level where price-discovery concerns bite -- so the
    float-threshold claim (price discovery survives) holds throughout.

 BOTTOM LINE: the structural-buyer endgame is bounded as the paper argues. The
 decumulation mechanism works, the ~10% plateau is a fair central estimate, and
 price discovery is preserved. The one amendment is to present the plateau as a
 ~6-21% range centered near 10%, duration-driven, rather than a single number --
 which if anything makes the bound more credible by showing its sensitivity.

