Very good sign that you (or rather, your system) have Daniilidou higher than Dokic. Some would probably laugh at Dementieva having better odds than Dokic, but that's of course where the draw sets in.
Instinctively, I would grant around 70% to Serena Williams for winning the title, as opposed to the 48.8% you operate with. If we ran this tournament 10 times, I would expect her to win 7 times. That may even be modest.
If we go by last year's form, she wins more than 60% of the tournaments she enters. Considering this is a Tier II (fewer rounds, weaker field than her average tournament) her winning percentage should be even higher.
What am I missing that the computer puts emphasis on? I sincerely believe that an oddsmaker, who goes by your chance distribution, will be vulnerable to heavy betting on Serena.