Favourite scales for implementing new character concepts

[the nerve-eater of Zur-en-Aarh]

I'm interested in what meta-level, as it were, people are most drawn to implementing new character concepts, on a scale ranging from new base classes with new mechanics (like oracle revelations or ki pools - is there a generic term for pools of resolve and panache and grit and so on?) to prestige classes to just making the concept distinct by a particular set of interlocking feats. In a context of "here is an interesting new sort of character to roleplay" rather than "this would let me minmax in new ways".

I'm particularly interested in reasoning behind, given that any new concept adds some more optional complexity to the game, which of those or other levels works best for you as most efficiently managing that complexity.

[Haywire build generator]

You have the priorities backwards. Some of us have statblock to backstory ratios of over 10 (Technically, I have yet to develop a full backstory unaided, so my ratio is undefined)

I'm just saying there's a lot more content for making distinct mechanics than there is for making distinct roleplay.

[the nerve-eater of Zur-en-Aarh]

Haywire build generator wrote:

You have the priorities backwards. Some of us have statblock to backstory ratios of over 10 (Technically, I have yet to develop a full backstory unaided, so my ratio is undefined).

To my mind the entire point of distinct mechanics is to support distinct characters.

I'm well aware plenty of people play with different priorities, and all joy to them, great that the game supports being enjoyed in many different ways; that's just not my cup of Ninepenny Knockout, and I was hoping I might reach some people in more harmony with my approach.

EDIT: On rereading it seems I was not entirely clear; this was intended as a question about people's preferences with regard to how new content is added to the game than with how people use existing content, though I suppose the two questions are pretty hard to separate.