validation
ModelledPlaced splice [M packaging]. CODATA-G evaluation [E algebra]. Baseline stays [P]. Derivation token unissued. External Hubble contact stays percent-class.G Prediction — Honest M13 Contact
External H₀ is the honest test. Planck gives δG/G≈−2.3%; SH0ES ≈+5.8%. The 0.04% figure recycles H₀≃69 from the same invert.
The published M13 invert relates to laboratory and a Hubble rate. It is a named package, not a derivation that avoids gravitational input.
M13 master equation for the gravitational constant
External H₀ contact (honest)
Insert an independent Hubble value and predict G (do not recycle 69):
- Planck km s⁻¹ Mpc⁻¹ →
- SH0ES km s⁻¹ Mpc⁻¹ →
The old 0.04% headline used a package-preferred — the invert of this same equation. That is circular (lab brick B-circ). It is not retired as a named dual; it is not evidence.
Evaluation on codata_2022 at CODATA G, named as K=c⁴/G [E algebra]. codata_2018 prints 68.9730537. The figure 68.97305 is that rounding. Baseline 68.973 and 3/8 stay [P]. One-convention Sciama+α duals to ~10.98, not 69.
Inputs of the published book
- — INPUT + [M] placement
- — laboratory
- — laboratory, when inverting for H₀
- — external test input (Planck / SH0ES), not a recycled 69
Dependence on the Hubble Constant
G tracks H₀ linearly in this book. Planck and SH0ES residuals are the Hubble tension, not an extra fail. The splice is placed [M packaging]. Its evaluation at CODATA G is [E algebra]. The clothes baseline stays [P]. The cross-book derivation token stays unissued. External Hubble contact stays percent-class. Do not retune bricks to chase 69.
What would make this a theorem (not claimed)
The coefficient is not waiting to be patched. Lab R-MCR-2 ruled out a one-functional derivation of the published k. Elevation to a free theorem would require a new principle, not a new number.
- External, tension-resolved H₀ inside the named band (gate G-H0).
- Percent-class G or m_e under that external H₀ (gate G-EXT).
- An ontology of R that is not naive expanding-R. If that radius expands, |Ġ/G|=H₀ is 928 times the quoted lunar-laser half-width. A stationary container does not write Ġ=0.
| Method | G () | Precision |
|---|---|---|
| M13 at recycled H₀≃69 | 6.674 | circular dual — not a test |
| CODATA 2018 | 6.6743 | 22 ppm |
| NIST-4 torsion balance | 6.6741 | 11.6 ppm |
Source Documents
The primary research documents underlying this page. All files are freely viewable and downloadable. They represent the working research as written — including integrity tags, equations, and finding IDs.
Documents open in the CPT Research Library — a formatted viewer for all source research files.