[rsd - edited to add "ability scores" to my list of modules]
SweeneyTodd said:
(For instance, the WHFRP system and setting both reflect PCs who don't grow in power to the point where they can do what high level D&D characters can.)
Power inflation is not a feature of D20. It is a feature of D&D. It would be very easy to make a D20 game where high level characters are not demigods.
I veiw D20 as a modular RPG system. These are the key modules:
1) The core mechanic (d20 + modifiers vs. target numbers) for task resolution
2) Six ability scores (STR, DEX, CON, WIS, INT, CHA; different names ok but undesirable)
3) States & Conditions
4) Skills
5) Feats
6) The game components (races, classes, spells, monsters, magic items, etc.)
I've ordered those in decreasing importance (from the perspective of the network externality). By the time you get to #5 and #6, I feel you can make extremely radical changes without introducing much in the way of inefficiencies.
A lot of people think that they need to start with D&D as the baseline, and then change D&D to whatever game they're targeting. I think that's the wrong approach. I think the best approach is to start at #1 and build up, making design choices at each step to reach a desired outcome. The result may appear to be highly divergent from D&D, but as long as it mechanically works the same way, and uses consistent definitions, the ability of a player to use the game based on prior knowledge will be quite high.
That's why I structured the original SRD the way I did and did not use the outline of the D&D core books.
If you want a low powered game with fragile PCs, D20 will accomodate you just fine. On the other hand, if you want a game where each different kind of task requires a different approach to its resolution mechanics, D20 is not your ideal platform.
D20 does not support linear character advancement well. Its minimum scale is 5% increments. If you want a game where little bits of the character are constantly changing in very minor, gradual ways over time, D20 is not a good fit.
D20 is not a good system for people who need to memorize all the rules. D20 assumes that you're using a rulebook as a part of game play, and thus a higher level of complexity is involved vs. what would be acceptable if the players were expected to have memorized everything.
D20 is not a good fit for large discrepencies in size & power. (i.e., mortals vs. gods, marines vs. battleships, bugs vs. people)
D20 is not a good fit for characters that change radically between (or in the middle of) game sessions (because of the amount of bookkeeping required).
D20 is not a good game for GMs who want to surprise players or use "hidden knowledge" of the mechanics as a storytelling feature.
But D20 is a good game for all kinds of genres that involve groups of people who band together and face challenges in return for increased power.