Ok, back on track. Time to think about conflict. Now, the reality is that if you're designing a game, conflict really means combat. Yes, absolutely, you might need to handle other types of conflict, but combat is the starting point. This is partly because it's such a central part of the hobby, but it's more because if your system is going to break, this is where it's going to happen. That sounds negative, so let me put it a little bit differently - no other part of a game system requires you to think about what the game will actually be like and how that interacts with expectations than combat.
That's a pretty loaded statement, so let me unpack it a bit into the four things that make combat a pain in the ass.
First, because combat tends to have explicit stakes (character well-being) that are measurable outside of the fiction (stats and information on the character sheet) its impact is directly observable in action. That is, while there may be some narrative interpretation applied to how 7 hit points got lost, it's still 7 hit points getting lost. This is not necessarily true of every system out there, but it's common enough to be a default.
Second, there's a huge range in expectations regarding how combat should go, so much so that it's one of the main yardsticks for judging genre. Games might have lethal, brutal combat, or they might have fast and loose, mostly harmless combat. Or they might have something in between. Whatever the case, the usual yardstick is that there is usually a specific look and feel the designer is going for, often epitomized by a particular book, movie or TV show though sometimes through their own personal lens of how fights work "in real life".
Third, fiction (both on-screen and in-mind) has the benefit of narrative protection for heroes, allowing an author to tell a tale of brutal, unforgiving violence without accidentally losing characters. A designer who seeks to emulate that may quickly discover that brutal and unforgiving cuts both ways. This can be addressed by effectively having different rules for players and everyone else (4E and many video games do this well) but that introduces a lot extra overhead into the system.
Last, we've got decades of training from D&D that have totally shaped our perspective. This is not a sleight to D&D but rather an assertion of its huge mindshare, as can easily be seen in most RPGs and computer games. And not just CRPGs - there are precious few shooters out there that don't have some form of hit points, and there's a good reason for that. If play is fun, then dying (and thus not playing) probably isn't. On the upside, this simple model has supported a lot of fun play, but it also means that the assumptions that might guide you in other parts of game design may fail you here. The blatant gamey-ness of combat does not stick out to people the same way it would in other parts of the game, and in fact if you move away from the established standards, people start reacting badly because they know how things should be.
Now, these are concerns, but not show-stoppers. But they're important to keep in mind as we move into designing conflicts.