As deployed by its proponents, MPI is insulated from disconfirmation: successful replications are read as entanglement correlations and failed replications as the predicted decline, so no feasible outcome of its own nominated experiments is treated as evidence against the model.

The insulating interpretive habit is quotable and real; the model's in-principle exposure (the NT axiom's universal prohibition, the matrix design as a positive prediction, untested fine structure) is also real. The claim is genuinely two-sided until MPI-R004 forces the question.

Proponents have demonstrably invoked the model's decline prediction to explain away the failure of their own pre-registered consensus replications, treating null results as a feature of the phenomenon. However, they also publicly conceded the paradigm is "not replicable," and the model contains universal prohibitions that render it falsifiable in principle. The strongest opposing consideration is that while the theory's fine-structure predictions could theoretically bind the model, proponents have not systematically pre-specified them in a way that risks disconfirmation in practice.

There is clear textual support for the charge that, in practice, proponents have treated failures as further confirmation: Walach et al. 2021 explicitly interpret their two null consensus‑protocol replications as “likely part of the systematic nature of such effects” (MPI‑E003), and the 2007 theory paper anticipates that the matrix design will never convince skeptics while presenting decline itself as a prediction (MPI‑E009, MPI‑E011). This lends weight to the claim that within the actual research discourse, both success and failure have been read as consistent with MPI. However, the model’s structure is not logically immune to disconfirmation: the no‑transmission axiom forbids any stable, exploitable psi signal, and the published theory makes specific fine‑structure claims (strength–reproducibility reciprocity, structured displacement) that could be bound into prospective tests. The strongest opposing consideration is therefore that the insulation observed to date arises from how proponents have chosen to interpret results and from a lack of prespecified, binding criteria, rather than from the theory being inherently unfalsifiable in principle.

The insulating interpretive habit is real and quotable: Walach et al. 2021 read their own preregistered failures as “likely part of the systematic nature of such effects,” which is heads-I-win, tails-I-win if left unconstrained. Against a blanket unfalsifiability charge, the model does expose a risky universal (any stable exploitable psi channel refutes the no-transmission axiom), it nominated the matrix surplus as a positive prediction that could have failed from the start, and the 2021 authors did conclude the paradigm is not replicable rather than simply declaring victory. The strongest opposing consideration is that this remaining exposure is mostly unused: the distinctive corollaries have never been prespecified as binding tests, so in practice the insulation is a habit of interpretation even though it is not a theorem of the formalism.

The interpretive-absorption pattern is real and quotable: the proponents' response to their own preregistered failures was that failure is "likely part of the systematic nature of such effects" — success confirms the correlations, failure confirms the elusiveness, and heads-and-tails both land theory-side up. But the claim's universality overreaches. The model has exposed flanks: the no-transmission axiom forbids any stable, exploitable psi channel anywhere, a risky universal that one replicable experiment would demolish; it nominated the matrix design as a positive prediction that could simply have kept failing; its 2007 paper contains distinctive, prespecifiable fine structure (reciprocity, displacement into related observables); and the 2021 authors did concede the paradigm is "not replicable" — a genuine concession about their chosen benchmark, not pure evasion. The strongest opposing consideration is thus that the insulation is a documented habit of interpretation rather than a theorem of the model: what is missing is not falsifiable content but binding, proponent-endorsed disconfirmation criteria, which the research agenda (MPI-R004) is designed to force.

The insulating move is documented verbatim: confronted with their own preregistered failures, the proponents wrote that the failure "is likely part of the systematic nature of such effects," which absorbs disconfirmation rather than registering it. But the universal quantifier in the claim is too strong for the record. The same team froze a consensus protocol on OSF, ran it, published the nulls, and conceded the paradigm is not replicable; Lucadou advised a field test with a directional, operationalisable prediction that came back null; and the NT axiom forbids any stable exploitable psi channel anywhere, a risky universal. The strongest opposing consideration is thus that the insulation is an interpretive habit rather than a structural feature — the model's fine-structure corollaries could be prespecified tomorrow. Both halves are true, which is what mixed is for; MPI-R004 would resolve it directly.

MPI-E013 slightly strengthens the model's in-principle testability side: its field prediction was operationalized and formally tested, even if the test was too underpowered to decide anything. Against that, MPI-E003's abstract still reads its own null as the predicted systematicity — the insulating habit remains quotable. The claim stays genuinely two-sided.

MPI-E013 shows a field prediction operationalized and formally tested, cutting against blanket unfalsifiability, and the no-transmission axiom carries genuine universal risk. Against that, MPI-E011 concedes in advance that no skeptic-proof replication is expected and MPI-E003's own abstract reads its null as the predicted systematicity — unfalsifiable in practice, falsifiable on paper.

MPI-E013 shows the model's field prediction was operationalized and formally tested, cutting against blanket unfalsifiability; MPI-E011 concedes in advance that no skeptic-proof replication is expected, and MPI-E003's abstract reads its own null as the predicted systematicity. Genuinely two-sided on the present record.

MPI-E013 shows a field prediction operationalized and formally tested, cutting against blanket unfalsifiability, and the no-transmission axiom is a genuine universal risk. Against that, MPI-E011 concedes in advance that no skeptic-proof replication is expected and MPI-E003's own abstract reads its null as predicted systematicity.

The model's proponents actively deploy the theory to interpret failed replications as confirmations of its core systematic evasion prediction (MPI-E003). This interpretive habit heavily insulates the core claims from standard empirical disconfirmation. The strongest opposing consideration is that the model technically exposes itself to falsification through specific, untested fine-structure predictions—such as the reciprocity of initial strength and decline rate—meaning the insulation is more a sociological habit of its defenders than a strict mathematical theorem.

Walach et al. 2021 explicitly interpret their own two failed consensus-protocol replications as “likely part of the systematic nature of such effects” while also labeling the paradigm “not a replicable paradigm,” illustrating a practice in which both success and failure are accommodated within MPI’s qualitative expectations (MPI-E003). This supports the charge that, as deployed, the model has often not treated any realistic outcome of its nominated experiments as disconfirmatory. The opposing consideration is that the NT axiom in principle forbids any stable, exploitable psi signal, MPI did nominate the matrix surplus as a positive prediction, and the theory articulates testable fine-structure (e.g., decline rate versus effect size) that could be bound in advance; this shows that MPI is not logically unfalsifiable, and the insulation stems more from current scientific practice than from the model’s formal structure.

Walach et al. 2021 read their own preregistered nulls as "likely part of the systematic nature of such effects," which is the insulation pattern stated as a claim about practice, and the 2007 design paper had already announced that the paradigm would never convince every sceptic. Against that, the model is not theorem-level unfalsifiable: the no-transmission axiom forbids any stable exploitable psi channel, the matrix surplus was a positive prediction that could have failed from the start, the 2021 authors did concede the paradigm is not replicable, and distinctive corollaries (strength–reproducibility reciprocity, related-observable displacement) could be bound tomorrow. The strongest opposing consideration is therefore that the insulation is a habit of interpretation rather than a structural feature of the theory, and habits can be closed by published disconfirmation criteria that this file does not yet contain.

The documented interpretive practice supports the charge in one direction: the proponents' response to their own frozen-protocol failures was that "the replication problem and this failed replication is likely part of the systematic nature of such effects," and the 2007 design paper pre-announced that no measure would ever convince every sceptic — heads and tails both read as confirmation. But the strong form of the claim, that no feasible outcome is treated as evidence against the model, is not established by this corpus: the same authors preregistered a consensus protocol, published their failures, and conceded the paradigm is "not a replicable paradigm," which is a genuine cost accepted; and the no-transmission axiom is a risky universal that any stable exploitable psi channel would refute. The strongest opposing consideration is that these concessions are about the paradigm rather than the model, leaving the theory itself untouched — which is exactly why the missing datum here is MPI-R004, an on-record elicitation of binding disconfirmation criteria. Mixed is a substantive finding: insulation is demonstrated as a habit of interpretation, not as a structural feature of the model.

MPI-E013 shows a field prediction operationalized and formally tested, cutting against blanket unfalsifiability, and the no-transmission axiom is a genuine universal risk. Against that, MPI-E011 concedes in advance that no skeptic-proof replication is expected and MPI-E003's own abstract reads its null as the predicted systematicity. GPT-5.1's framing — unfalsifiable in practice, falsifiable on paper — is exactly this two-sidedness.

MPI-E013 shows a field prediction operationalized and formally tested, cutting against blanket unfalsifiability, and the no-transmission axiom carries genuine universal risk. Against that, MPI-E011 concedes in advance that no skeptic-proof replication is expected, and MPI-E003's own abstract reads its null as the predicted systematicity.

The pattern is real and quotable: the proponents' response to their own consensus-protocol failures was that failure is "likely part of the systematic nature of such effects." But the charge is not airtight — the model does expose itself at the aggregate level (a stable exploitable psi signal anywhere would refute the NT axiom), it nominated the matrix design as a positive prediction that could have kept failing from the start, and the 2021 authors did conclude the paradigm is "not replicable," which is a concession, not an evasion.