I'm not sure if you're the right person to ask, but you're a good starting point at least.
I'm looking to reverse-engineer a few rules to get a feel for the underlying assumptions, expectations, and targets... both so I can have a firmer sense of how to adjust encounters for my parties, and also for some general game design.
Is it fair to say that the stat blocks available for the iconic characters are roughly representative of the power level Paizo adventure and monster designers build for?
(This will help me reverse engineer things like expected chance to hit with primary attack, or how likely a monster is to make a save. Good both for figuring out how much to beef up the monsters in my games, and for non-Pathfinder design.)
No. The stat blocks for the iconics are variable. Some versions are meant to be easy to understand and standardized for ease of play in the Pathfinder Society. Some versions are meant to be customized for earlier Adventure Paths. Some versions are meant to simply portray them as the characters they are, personality wise. They are NOT meant to be anything close to "optimized" characters for their class, and in fact, most of them deliberately make choices that optimizers gnash their teeth at.
The BEST way to get a firmer sense on how to adjust encounters for your players is NOT to reverse engineer other people's work, but to pay attention to how your OWN design works and interacts with your group, and constantly adjust things to compensate for their own strengths and weaknesses.
When we build encounters and monsters, we generally assume a middle-level of player experience and a 15 point buy for a character with a party of 4 or MAYBE 5 characters. For Pathfinder Society adventures, we assume a group of 6 with 20 point buy and a LOT higher player skill... or at least, that's what we SHOULD assume.
Table 1–1 in the first appendix of the Bestiary is what we assume for building monsters, though.
I'm aware iconics are not optimized, and I'm quite fine that they're not. My goal was really to have a better sense of the PC equivalent of Bestiary Table 1-1; I know my PCs are more powerful than the adventure path assumes, but I was looking to get a somewhat more quantitative assessment of how much more powerful than the current qualitative assessment. That would, I think, make it easier for me to adapt existing published content to my game; providing a baseline for the adjustments to start from, at least.
I'm adjusting, but I'm still developing a feel for how much is suitable, and I was looking for a more... scientific approach, to treat it a bit less like an art form, at least until I get more comfortable.
That, and to unravel some of the underlying assumptions of Pathfinder for the sake of understanding them and deciding how I want to translate (or not) those assumptions to the system I'm oh-so-slowly designing.