There is only one principle: self-reinforcing gravitational feedback. Everything else — cylindrical flux geometry, flat rotation curves, the Tully-Fisher relation, cosmic filaments, lensing amplification, the Bullet Cluster mass offset, and the apparent need for dark matter — are its derived consequences. Previous 2 versions presented flux geometry and feedback as two parallel postulates. This was logically incomplete. Without feedback driving matter into disk structures, there is no cylindrical flux geometry. The two are not parallel; one causes the other. Version 3. 0 corrects this by establishing feedback as the sole foundational principle and deriving everything else from it. The causal chain: Initial density fluctuation → potential gradient → flux concentrates along gradient → matter accumulates → disk/filament forms → flux geometry changes (Aₑff ≠ 4πr²) → gravity enhanced → flat rotation curve → expressed in spherical language → NFW "dark matter" profile. Five derived consequences, all parameter-free: Flux geometry — Feedback drives matter into disks; disks confine flux cylindrically; Aₑff = 4πHr replaces 4πr². The constancy of H is verified numerically from the Miyamoto-Nagai potential (4. 2% variation across galactic radii). Dark matter as mathematical artifact — The "missing gravity" Δg between cylindrical and spherical propagation, when attributed to an unknown mass distribution via Poisson's equation, yields ρDM ∝ 1/r² — the NFW halo profile. Dark matter is not a substance; it is the mathematical shadow produced when feedback-governed, cylindrically-propagating gravity is expressed in the language of spherical propagation. Cosmic web directionality — Isotropic (spherical) gravity has no mechanism to produce filaments from a uniform distribution. Feedback does: it selects the direction of maximum potential gradient, deepens it into a 1D flux channel, and drives the inevitable formation of filaments as stable attractors. The cosmic web is not a numerical coincidence — it is the large-scale fixed point of feedback dynamics. Gravitational lensing — Disk-plane lensing is amplified by r/H ≈ 33× for a typical spiral galaxy, with a viewing-angle anisotropy (edge-on vs. face-on) that discriminates this framework from dark matter models and is testable with Euclid. At cluster scales, Void-Anchored Lensing (VAL) explains the Bullet Cluster mass offset (~150 kpc) as a consequence of filament flux nodes following galaxy positions through the collision, without dark matter. Feedback threshold — fcrit = H/R (disk aspect ratio) defines the boundary between feedback-active and feedback-free regimes, verified across all major astrophysical system types. This threshold also defines the domain of validity of Newton's inverse-square law.
JongJin Ma (Wed,) studied this question.