check parameter PCR_PROBABILITY

For the simulation of the Illumina Body Map, I get from ~5M reads to ~50M fragments, without filtering and PCR. PCR with a certain PCR_PROBABILITY (p) parameter then leads to a number of fragments :

p=0.1, n= 3,908M
p=0.05, n= 1,330M
p=0.0001, n= 1,342M
p=0.00001, n= 1,299M

To me the value obtained for n seems a bit high overall, and also there is no notable difference between p=0.05 and p=0.00001.