Two reviews are not a vote
Final compared the arguments, not the count.
C1 supplied a valid counterexample. C2 preferred a narrower operating rule. Final retained the counterexample because C2 did not disprove it.
Decision trail
Follow one technical claim from origin to unresolved decision.
MP Fusion preserves reported output lineage so a reader can see why a claim appeared, how it changed, what survived, and what evidence could change the conclusion later.
Photocentric · oil-family separation
A real correction, reconstructed from the reported outputs.
A3 / B-02a → S-06 → C1-R3 and C2-OIL → Final → Residue.
- A3 / B-02a
Original route options
Discovery kept whole-oil conversion and selected-family separation available as different, conditional approaches.
- S-06
Restrictive Synthesis gate
Synthesis limited fractionation to cases where whole-oil endpoint control could not repair the variability.
- C1-R3
Economic counterexample
The technical review found that whole-oil conversion could work yet cost more to correct downstream than a fractionated route.
- C2-OIL
Different narrower view
The operational review preferred the narrower trigger but supplied no evidence that eliminated C1's economic counterexample.
- Final
Restored comparison
Final restored the broader comparison and retained whole-oil control as a fair baseline. It did not claim a profitable fractionation process.
- Residue
What could decide it
The unresolved decision needs actual whole-feed function, recovery, reagent use, wanted-material loss, and total processing economics.
What the trace establishes
A reason for the changed answer—not a profitable route.
The record explains why Final differs from Synthesis and what data remain necessary. No conforming material, cost advantage, or profitable process was demonstrated.
What the trail may show
A reported witness to the investigation.
- 01
starting objective and authorized evidence boundary
- 02
frontier foundation and separate Discovery contributions
- 03
relationships developed during Synthesis
- 04
Challenge arguments and material claim changes
- 05
surviving directions, challenged-out claims, and useful Residue
- 06
unresolved questions, reconsideration conditions, and next discriminators
Provenance boundary
Output lineage—not security-enforced telemetry.
The trail reconstructs reported stage outputs, source references, claim changes, and candidate fates. That makes a decision easier to inspect; it does not make the decision true by itself.
- It is not permission enforcement, runtime authorization, or a security certification.
- It does not expose private model thoughts or attest hidden provider internals.
- It does not replace experiments, field validation, or customer-specific economics.