Store expected number of signal events in sensitivity

[maxnoe]

After #90, we should also store the expected number for 50h in the output sensitivity table, not the number of weighted mc events.