The integrated MadGraph4 cross section for the trident file was 0.14566 barns, which translates into 1166 trident events per 500k beam buckets, so this agrees.

The number also matches how many tridents were sampled before the actual mixing occurred, so none were lost in the process.

Similarly, 4977 WAB electrons were sampled and mixed, agreeing with the calculated number of 5047 (0.63059 barns).

To get these numbers:

#events per 500k bunches = 

Sigma*(6.306*^-2 atoms/barm-cm)*(0.0004062 cm)*(625 electrons/bunch)*(500000 bunches)

So this shows that the MadGraph cross sections can be recovered after mixing.

Whether or not these generated cross sections are accurate is another question.