Here are some interesting stats about the accounts in NEMESIS:
avg post count: 211
avg registered days after NEM start: -64 (meaning registered 64 days before NEM start on avg)
avg number of days since last active (from 3/8): 65
For comparison, here are the same stats for sock accounts:
avg sock post count: 5
avg sock registered days after NEM start: 11
avg number of days since socks were last active: 203
You can't rely on this because Ripple Giveaway led to a lot of sockpuppet accounts created in advance (because of expectations that another coin will use the same approach). For a better assessment you should do Resource Testing trick which would show how many computers are owned by the remaining accounts. If you get that only 30% accounts own a computer then you could reliably assume that 70% are sockpuppets.
In this case, post count is a fair approximation for a proof of work

People who have only 1 post in the past year are not likely to be real people. And if they are, then they are even less likely to be able to contribute anything to NEM.