I believe the correct algorithm for determining best overall should as
follows:
For each tournament, assign the number X / (number of entries), where
X is your place finish, as a score.
Multiply all the scores for a given individual together.
The person with the smallest score is best overall.
In other words, the events are independent events. The a priori
probability of an entrant ending up in position X or better is X/N.
Whomever achieves the joint result with the lowest probability has done the
best.