Sort of a side story but hopefully illustrative of the problems in this space. A couple of months ago I started on a side project to do a game creation suite modelled on the Dvorak card game (a Nomic varient). This is a very small subset of what y'all are trying to do, but some similar problems.
The very brief summary of the game is that Dvorak is a very basic card game with each card created at the beginning of the game. In other words, custom game creation within limited parameters. If you want more on the ruleset google and go to the co.uk site. My idea was to take this in a 'Magic the Gathering' esque direction. Set up a basic model for combat, and have each card be taggable. Each tag would contain a function as a String (some extension of Rails's acts as taggable on steroids plugin and Raganwald's work on methods as Strings in Ruby). Consequently, any action could be custom defined in ruby code as a method and then attached to a card. The end user might not be proficient with Ruby but could create their own card set with the attributes/tags provided. I also had a basic attack/defend hitpoint system designed , never implemented.
Well, the short story is I found that from a design side this worked great. In fact, I used the same principles to make several card games and played them with friends locally (in Tokyo). But translating this concept into computer code was far too difficult and time consuming and with little or no upside. Doing anything even marginally visually satisfying w/ the GUI (esp. off Rails) is also no small undertaking.
Consequently, my advice in this space is to:
(1) Start small - single concept which could be expanded upon and preferably has some inherent profitability
(2) Extensively discuss the technical details w/ someone who knows their 'shit.' No offense poppy, but you don't seem to have done this and much of what you say seems impracticable.