Estimate of TRV matrix elements.

First, the elementary observation that measured Mf/Mgt is only varying
by 1 order of magnitude for 4 orders of Mgt changes. Intuitively, as nuclei
become more complex with A and GT strength becomes more fragmented, the antianalog and its isospin mixing producing Mf become similarly fragmented.

Given the standard schematic wavefunction for the analog

the antianalog has opposite sign to make it orthogonal.

Writing the isovector part of Herzceg's general TOPE nucleon-nucleon potential,


one can immediately see this will flip the relative sign of the antianalog and
produce greater overlap with the analog. For both long-range and contact TOPE
interactions this will make large matrix elements with the analog.

The figure of merit then
<A| Vtrv | Abar>/<A | Vcoul | Abar>
becomes for clean antianalog states nearly constant. BBS's calculation of
large <A|Vtrv|Abar> in 134Cs appears to us to assume Abar is clean, that
an experiment could sum over all final states of given Ipi and thus sum
over the antianalog configuration's isospin mixing strength.
This is a reasonable experimental assumption for D as long as Qbeta is not changing much, and could be done by summing over gamma transitions for E.

Any figure of merit increase for Vcoul small appears qualitatively illusory, as selecting a final state for small Vcoul will naturally suppress Vtrv as well.