Regarding the misattributed paternity numbers/rates:
Obviously interviewing people isn't going to reveal the correct numbers: If a mother isn't going to tell her child, or doesn't know [switched in the maternity ward, chimera, multiple partners] chances are they aren't going to tell some random researcher either.
However, there was some interesting research at a hospital in the UK (maybe Bristol ?) which tested the men which were present at the birth of the child against the child, and the numbers are rather high - something above 1 in 20. And this is for the men that were present.
Something which should be more common knowledge.