You Surprise, Enemy, You, Enemy, You, Enemy, and on and on...
or
You, Enemy, You, Enemy, You, Enemy, and on and on...
hmm ...
(1) - Surprise You , Init check win , You , Them , You , Them ...
(2) - Surprise You , Init check lose , Them , You , Them , You ...
(3) - Surprise Delay , no init check , You , Them , You , Them ...
So taking the Surprise round, means you get 1 action, and then variability kicks in, and that might not be good.
Not taking the Surprise round, means you get Full action, and less variability, and that might be better.
I think in Optimizing, reducing variables is almost always the better choice, but certainly not always.
EDIT - oops, bad day, forget the 1/dayDivine Oracle PP for Clerics, jacks up the surprise round, and gives 2 initiative checks 1/day.
Even getting surprised by the bbeg, this yields:
(4) - No Surprise , Init check semi-auto-win , You , Them , You , Them ...
(4a) - No Surprise , Init check win 1/day , You , Them , You , Them ...
I'm assuming semi-auto-win because, a CO party is gonna have at least 1 good initiative PC, and 5:<5 fights favor the Party.
(4a) gets a win because the reroll init check works out to +4 to +5 equivalent, most of the time. Solos are in trouble.