The files 

ene_q2_num_event_le_neug.dat
ene_q2_num_event_me_neug.dat

are in bins of Q2 and energy and have the same
format at the previous files.
ene_q2_num_event_le.dat 
ene_q2_num_event_me.dat 

The are in the same area 
http://www.pas.rochester.edu/~hbudd/qe

  The nuclear correction is different. I used the Fermi
gas correction from Neugen for carbon. Hugh sent me 
the routines a couple of years ago. Back then, I put them in.
A plot of the correction is fermi_gas_cor.eps .

   The origin of the integration problem when
the limits  in Q2 is set to the kinematic limits is understood.
   The 1-d integration starts by averaging the endpoints. Next,
it doubles by a factor 2 and sums. It checks that
 % abs diff((step i+1) - (step i)) < Integration Accuracy.
The 0.001 is 0.1% accuracy.
It continues to double the number of points until
Integration Accuacy is achieved or it hits the max number
of step. If too many steps, it prints out an error and returns
the last value calculated.

  When the bin limits are set to the kinematic limits. 
The 1d energy integration at the 2 Q2 endpoints have only one 
non-zero point. Hence, the integration for those two
lines can never converge.  I set the size of the bin slightly
smaller in Q2. The iteration, ene_q2_num_event_me,  had Q2 limits set from
the center of the energy bin.  ene_q2_num_event_me_neug.dat
has the Q2 limits set using (upper energy bin -0.001).
ene_q2_num_event_me_neug.dat integrates over almost the full bin 
while  ene_q2_num_event_me.dat doesn't.