Attachment 1 shows a comparison of the zero-degree telescope from the decay measurement with the 26gAl beam produced today (left side) and a GEANT4 simulation by Shimizu kun 
some months ago.  

I discovered two reasons that July 2016's beta spectrum did not match the simuation:
    1) The threshold was too high, so we should try to decrease it during another test soon
    2) There was a strange bug in the code that also affected the spectral shape.

Now the comparison of the beta spectrum between our measurement today and the GEANT4 simulation is excellent.  (There are some differences in the setup, so that may account for 
some of the small variations).

My calculation of the isomeric purity output the following results:

root [0] .x macros/purity.C 
Run number: 46
Solid angle: 0.0386799
Betas: 3.42938e+07
Purity (PPACa): 0.862612
Purity (PPACb): 0.906841
Isomeric purity (PPACa): 0.281095
Isomeric purity (PPACb): 0.276141

(This differs from the online logbook, as I mistakenly used the PSD solid angle instead of the first SSD layer.)

Keep in mind that the lowest-energy tail of the beta-spectrum in the experimental data are absent, and so the actual isomeric purity is slightly higher than above (but it seems 
negligible if I take the 25 keV/bin setting, assumed from 0~200 keV misses around 500 counts per bin)

Still, the beam seems to be around 70% ground state.  However, keep in mind that our isomeric purity from the test run was somewhere around 42% so the difference is only 10% 
which may be small, and so we should make a stronger effort to make the higher purity isomeric beam, maybe tomorrow.