In Burning Wheel, if a player succeeds at a test, intent and task are both realised. All the GM can do is embellish the success.
If a player fails, then the GM narrates a consequence. This is either implicit in the fiction (analogous to "make a move that follows) or else has been expressly flagged in advance. And it must negate the intent of the test.
Because the test will be in the context of a scene that has been framed in accordance with the rules that I already stated above, it will put at stake something that pertains to or bears upon player-authored priorities for their PC. And so the intent will likewise bear upon or manifest those priorities in some fashion. Hence, the consequence - in negating the intent - will also bear upon or speak to those priorities. So it's not only
make a move that follows the fictional play but
make a move that follows the thematic play.
So the GM's decision-making is very focused. This is what makes Burning Wheel more "intense" and personal than more conventional, adventure-oriented FRPGing.
In MHRP, actions are always opposed - either by an opposing character, or by the Doom Pool - and failure can mean "no effect" or may involve the GM spending a Doom Pool die to impose an effect on the character whose player failed. If a scene has stakes beyond defeating the opponents in it, these will be manifested as a Scene Distinction (eg Vulnerable Bystanders), and if the Scene Distinction is still there at the end of the scene, the GM gets to bring home whatever would follow (eg the heroes failed to protect the vulnerable bystanders).
Because all consequences are mechanically defined (as Scene Distinctions, or as dice-rated Effects), they are rather more stylised than in Burning Wheel. I would say that, overall, the GM has less scope to put truly
hard choices and situations in front of the PCs. In my experience this makes the game less intense, on the whole: not every roll feels like it might be
the end of a character's hopes and dreams. Whereas Burning Wheel leans more in that direction.
I personally don't recognise these games in some of the descriptions, in this thread, of "narrativist" vs "trad" RPGs. In particular, my experience is that in trad-sim-immersionist RPGing,
what is at stake is often quite opaque to the players. The GM might be constrained by pre-decision (prep/notes/map/kehy/etc) but the players frequently don't know what the content of these constraints is.
The upshot in play, at least in my experience and also what I observe/hear, is that players often try to make low stakes action declarations to try and learn more about the situation, before they fully commit. In dungeon-crawl play, this is exploring, listening, mapping etc without opening doors; in mystery-solving play it the collation of documents, rumours etc before confronting suspects/cultists etc.
Another manifestation of this
start with low stakes idea is that play should begin at the "edge" of the real action (eg the entrance to the dungeon; the city gate; the edge of the wilderness). When I've posted about
the first Burning Wheel session that I GMed, in which I began the action in a bazaar with a peddler offering an angel feather for sale (which spoke directly to one of the PC's Beliefs), more than one poster has said that I "should" have started play at the entrance to the city, so the player could make his own choices about how (as his PC) to find his way to the bazaar. This is about starting with very low stakes ("Do I successfully enter the city?" Almost certainly,
yes), and having the player engage in low-stakes action declarations (looking around, talking to bystanders) before the high-stakes opportunity to acquire the angel feather is established.
Speaking purely for myself, I have come to prefer RPGs where the stakes are reasonably clear, and players are not engaging in low-stakes action declarations just to work out what is going on and find a way for their PCs to get to the higher-stakes action.