Figures from Ben's thesis:
7.16: The Compton edge from the 511's is large in 37K decay, even in
coincidence with the eMPC. This determines the energy threshold we used,
330 keV in the plastic scintillator.
   47K isospin breaking will be constant with Ebeta.

The effect of the 1.8% 37K branch making 2.7 MeV gammas is shown in simulation.

7.17 is mostly details of 7.16. For the precision measurement asymmetries shown
later, there is spectrum distortion from the higher-energy gammas above
the threshold.

7.19 shows there are still small difficulties with the delta-E coincidence, but
above threshold they are reduced to manageable size.


8.4 suggests most wall backgrounds go away with the eMCP, but the delta-E is
needed to get rid of gammas from the trapped atoms.

Note from the 47K level diagram:
There are two large branches making gammas of up to 586 keV that will resemble the annihilation radiation effects. An energy threshold of 408 keV Compton edge
(+10% for resolution?) will help a lot with those.

The 2.0 MeV gamma will have 50x the branch  of the 37K 2.7 MeV gamma.

Yet the beta spectrum between 2 and 4.6 MeV will be pretty clean of gammas,
and have more than half the useful beta-. 
A correction for the higher-energy gammas (see also J. Smith's publication)
might be possible since the Compton edge of the 2.01 MeV will likely be
visible to singles, so one could scale from that.

So it's likely that for things like a check of polarization, and maybe consistency of isospin breaking, even without a delta-E a 47K plastic scintillator is viablea and worth doing. It would certainly help a lot in singles to suppress
gamma-ray backgrounds. 
