In the attachment you find (zipped archive of) a collection of those
events. Each event is in a separate file, each line corresponds to a
tower/particle with px in the first row, py in the second, pz in the
third, and E in the last row.

The 'FatJets.*' files contain events with jets that have particles/towers
more then 4*Rcone away from the jet axis. The jets are reconstructed using
the fixed MidPoint algorithm (R' = Rcone/2, Rcone = 0.7) with an overlap
threshold of 0.5. I picked 2 events from CDF data (Jet100), 7 events from
a Herwig sample with pThat = 100...150 GeV/c (fully simulated), and 3
events from the same Herwig sample at the hadron level. The total energy
of all particles should be 1960 GeV in the last three events. I am not
sure if I verified this, but if there is energy missing, then this is due
to some particles failing the |eta| < 7.5 cut that we have to apply inside
the CDF software. (We use single precision variables only, so the
calculation of the rapidity of particles that close to the beam pipe can
become problematic (log(0)).)

The 'DarkTowers.*' files contain events with a significant amount of
energy (>~ 50 GeV) missed by the unfixed MidPoint algorithm. There are 10
events from CDF data (Jet100), 11 events from the pthat = 100...150 GeV/c
Herwig sample (calorimeter level), and 17 events at the hadron level.

I selected these events by looking at the jet configurations computed
within the CDF software, but I checked that I obtain the same
results/effects when my stand-alone algorithms are used.

I just noticed that the full Herwig event histories of the hadron level
events might also be interesting. I can provide them if someone would like
to have them.