Many years ago, I played Spycraft. I loved the first edition, and when Spycraft 2.0 came out, I snapped it up. And promptly never played Spycraft again, and not because I stopped loving the idea.
See, there were somewhere around a hundred little dials the GM could tweak for any given campaign, and many of them sounded awesome. I was paralyzed by choice — I suddenly wanted to run five different games, and consequently I ran none. Similarly, whenever I see a game boast something like “over twenty classes!” in a core book, I shake my head at it.
Here’s the issue: as a designer, you know the optimal choices, the impacts of various choices, and the interactions of compound choices. But that’s you; people new to your game have none of that. Reading all of those options can be overwhelming to some people, because now they’re playing the game of “make optimal decisions.” Others can get inspired with a host of ideas, that can be incompatible with the host of ideas that other players come up with — too many options and choices leads to an inconsistent world.
So, if you end up piling more and more into some part of your game, slow your roll. Strip some of it back. That will help your game get traction as you don’t have such a hurdle of overwhelming information. Pick your favorite half or two-thirds of that content, and hand to the reader what you really love about your game. Play with that — there’s certainly no science to this.
Now, nothing says what you strip away needs to be deleted. If your game takes off, you have more content to hand out, and as people gain mastery of your game, the new sets of choices are not as overwhelming as giving them all at once.