I like your idea and think I would enjoy playing in your game. How does this scenario fit into the problem threat resource reward theory?
The party has had some conflicts with the thieves guild in town. One rogue from the guild approaches the party cleric saying he wishes to change his ways and try to become a follower of the cleric's deity.
In this situation, there is no immediate threat and there are both problems and rewards for different courses of action that the party might take.
If they take him in, they may learn he needs protection from an enemy he recently made in the guild. The party may become a target if they protect him from assasination and they may have to keep him at the temple which may also become a target. Also, they will find that he is serious about changing his ways, but still has links to the seedier elements of the city and falls to weakness sometimes (occurring gambling debts that he can not pay). However, the reward will be an informant that knows the city well and will tell them everything he can. They'll have additional information at the beginning of every quest that starts in the city. So along this path they have risks and rewards.
If they don't take him in he will need to flee the city and end up hiding out in a village where he could either have to continue a life of crime to get by (and turn into a BBEG for them to fight someday) or find a way to go straight on his own (and someday become an ally but perhaps one that is bitter towards the party). The party will not have information for their conflicts with the guild in the near future. The party will not have the focused attention of an angry local theives guild for protecting someone they are trying to off.
The difference here is that there is no real problem and no immediate threat yet the players do have a chance to shape the way the story will go.