But at the same time, show the value of those answers, because on the face of it, knowing how much is 342x520+92 is not really valuable. Even knowing how much is any number plus any number is not that interesting for basically 90% of the population.
It's not clear to me that there's an obvious right answer here, but I do believe that most obvious answers are, to some degree, wrong. Namely, saying "just ask harder questions" doesn't take into account that then the student is trying to solve hard questions and learn a new method at the same time.
Teaching a more general method on easy problems gives time to become acclimatised to the method, while also offering sanity checks (by easier, faster, and / or approximate methods) to make sure that it's not going wrong. On the other hand, as you say, this approach seems unmotivated, and bores or frustrates students—there's no obvious right answer! I think that all that can be said is that one must find the appropriate balance for each individual student, or class.
In an ideal world maybe the teacher could tweak the problem each time they see a person using a different method, to make the problem one that forces a person to use the method they are trying to teach. In the real world you usually can't give each student that kind of attention. You just have to say, "Look, I know you could solve it that way, but I need you to try to solve it this way so we can learn this method."