gstlal-inspiral/python/far.py: increase singles penalty

There is a troubling L1 single that keeps coming up in the MDC, event page: G910434. The qscan for L1 clearly has glitches in it, and if this kind of event is found in O4 it will likely become a retraction. We want to reduce this possibility so we think the best thing to do given time constraints is to increase the singles penalty. Here is a plot of LR vs FAR of coincs in MDC11:

This shows that if we just increase the singles penalty by 1, the retraction (LR ~ 15.7, FAR just below 2/year) will move above the 2/year threshold.