[MENTION=6688277]Sadras[/MENTION], I read your post as setting out an outline for prep, not a sketch of the form that play might take. If I got that wrong, sorry. If I got that right, read on!
I think that any pre-set sequence of events for the game creates "Mother May I" or similar sorts of issues - ie the GM has already decided how things will go, and deviating from that is at the GM's sole discretion.
I may have misunderstood [MENTION=85870]innerdude[/MENTION]'s post, but I took it that he is looking for a method of establishing and signalling consequences as part of the setting up of situations in the game and then resolving the actions that the players declare for their PCs. To me, this is a
system problem (using a fairly liberal conception of "system"), but not a planning problem.
I'm focusing especially on this from the OP:
I don't want to turn the game into a game of escalating consequences, for which the player(s) have no recourse other than to cow-tow to what I'm presenting. I want them to have avenues for success, while still balancing the need to present challenges.
So how do I do this better? How do I introduce consequences/complications that are A) interesting, B) have real dramatic heft within the fiction, and C) don't require the party to start finagling with me as the GM?
As I read it, the worry about "kow-towing" to the GM and its close neighbour (c) "finagling with the GM" (= Mother May I, I think) is this: if consequences are set which (i) have heft, but (ii) aren't clearly integrated with player-side action resolution mechanics, then the players (iii) will want to avoid them (because of (i)) but (iv) won't see any player-side/mechanical-type way to do that (because of (ii)). Hence the game will degenerate into kow-towing and Mother May I as the players try and avoid the consequences in the only way that seems possible to them.
If I've got this right, then the solution is to more clearly frame action declaration and the surrounding context in a way that both makes the consequences clear to the players, and makes it clear how they're an outgrowth of play rather than a GM's arbitrary stipulation. DungeonWorld provides a clear model for this that is fairly well-known on these boards; but I'll got to Burning Wheel instead because personally I know it better.
BW actually offers two approaches: the official one; and the one the designer actually uses, which he discusses in the designer notes section of the Adventure Burner.
Official BW: on the player side, an action declaration involves stating task (ie what am I (as my PC) doing) and intent (ie what am I (as my PC) hoping to achieve, given the context of my action declaration). (Note how this contrasts very much with a certain type of classic D&D/wargame-y type play where players delcare tasks while keeping the intent secret from the GM, hoping thereby to establish buffers against GM-narrated consequences.)
On the GM side, an action declaration requires making sure the player is clear about the surrounding fiction, so that intent and task make sense, and so that the appropriate skill can be determined; requires establishing the skill to be checked; and requires making clear to the player what the consequence of failure will be. The GM, in doing the last thing, is advised to focus more on the intent then the task, so that the result of failure will be that the player (and PC) doesn't get what s/he wanted to out of the situation.
The GM has to make the consequences clear before the dice are rolled, so the player has the chance to use player side resources to boos the chances of success that seems warranted to him/her in light of what is at stake.
BW as played by Luke Crane: on the player side it's the same. But on the GM side, the first and third steps are merged: that is, the consequences of failure are taken to be implicit in the GM's narration of the context for the action declaration, which should be drawing heavily on the prior play of the game. This drifts BW closer to DW as far as this particular aspect of play techniques is concerned.
On this approach, the player has to make the decision about committing resources without having any definitive statement of what is at stake, and instead depending on a shared intuition with the GM about how the current situation and its imnplicit stakes fit within the broader and unfolding content and trajectory of play.
Notice how, on this unnoficial approach, if the player and GM come apart in their understanding of the context and trajectory of play, then the player might feel blindsided and/or railroaded by narrated consequences of failure. Which starts to push towards the Mother May I problem [MENTION=85870]innerdude[/MENTION] wants to avoid. And conversely, if I've understood innerdude's problem right, the first step is to improve the degree to which the GM is communicating what is at stake in the way that situations are framed and action declarations resolved.
(For what it's worth, I mosty use the second, non-official approach when I GM Burning Wheel.)
A second thing that has just occurred to me is this: both BW approaches take for granted that once a sittuation is properly described by the GM, and once a task is properly described by a player, the correct mechanical approach will be evident. If that part of action declaration is breaking down, though, then Mother May I problems can arise.
A simple example: if the situation in an AD&D game is that the gang leader wants to kill the hostages, but its known he will be generous if he is given a nice cake for his birthday, then we are heading in a Mother May I direction. Because AD&D has no rules to resolve a player declaring that his/her PC makes a nice birthday cake; and has no rules to resolve a player declaring that his/her PC goes out and buys a nice birthday cake.
So as well as thinking about how framing is conveyed, innerdude may also want to think about how actions and consequences in the situations being presented to the players map onto the range of mechanical options that are available to the players, given the RPG the group is playing.