The Interaction Designer's Dilemma

As an interaction designer I often face a dilemma… which path to follow… Interaction design is by now an established discipline. An established discipline means that a lot of intellectual jargon has got associated with it. Related think tanks of the subjects have already devised a set of guidelines and processes to follow and emulate.
Here comes the problem…
Designing for interaction unlike other subjects is an extremely fluid field. The designer has to face a variety of problems. And not problems may be solved through fixed categories of contextual enquiry, usability goals, heuristic analysis etc. That’s jargon, which comes useful when a person is designing a ‘user centric’ solution to a problem. But things don’t always work like that. True, interaction design should keep the human factors paramount. But sometimes to design for human factors and user centric approach fails in the face of logical problem solving. Problems are not always independent entities. They are often part of an entire process or set of events. Sometimes the best solution is not just designing for the user at each node of the process but to try and achieve an overall optimization of the process.
So what path to follow for this? What can be an ideal approach to solve such problems? Clearly the standard HCI design principles miserably fail in this context. Then what is the ideal successful problem solving approach.
Truth is, there is none. So right now all that is possible is to adopt a hybrid approach which keeps the standard principles in mind but formulates its own way through logical reasoning and careful analysis of the entire process. Subsequently identifying the key points where various users interact with the bigger picture and find out how those interactions can be made useful and more fruitful.