I encountered Spelunky last week, and I notice that the creator's words for it radiate deeply:
"My goal was to create a fast-paced platform game that had the kind of tension, re-playability, and variety of a roguelike. In roguelikes, the gameplay tells the story, and I wanted to give Spelunky that type of a feeling..." (Source)I think that deep down this is what I'm trying to make: a roguelike inspired with a bit of a System Shock theme but with virtual worldly multiplayer support. The trouble is that there's so many different ways to do this!
- Do I go with complex, thought-provoking economy or a simple-but-fun economy?
- Do I go with a virtual world without an apparent end, a session-based game with an end?
- Do I allow the players to change the world directly, or implement a more realistic indirect world alteration mechanic?
Clearly, I'm operating well outside my design document I ironed out last month. I suspect I could make some really good progress if I could settle on my new one. Thus, that is my current focus: making some solid decisions along these lines.
However, I personally feel that it would be more enjoyable to have infinite areas to explore - areas with real detail and things that are fun to discover - than to have a world that you can change dramatically.
Oh and the best part of Spelunky is attempting to blow up the shop keepers without getting killed. Then you can loot their shops. :)
(Total side point: the quote, links and bullet points make this post MUCH easier to follow than some of your others.)