Your view that the user is never wrong is to some extent not only a shared sentiment but a fundamental UX concept. Here is some food for thoughts.
There is no such thing as stupid users - only expert supremacy
People often feel 'stupid' due to lack of experience or expertise. This is normal behaviour - if you travel to Enfield and buy an overpriced match tickets from a tout, only to find out at the gates that they are fake, you, your friends and locals (who may have done the same in the past) may consider the act 'stupid'. But it's rather a case of inexperience, not stupidity. If you repeat the same mistake twice, you are a slow-learner, but not stupid.
It's not the designer's fault
User testing is integral in UX to confirm assumptions made by the designer and to explore unforeseen issues. If only designers could test a system before it was designed; but per Einstein, this would violate the cause-and-effect principle.
Users to designers are like toddlers to adults
We don't blame, or call stupid, a 2 year old toddler for trying to drink from a sealed bottle - we acknowledge they haven't mastered many skills yet. Adults, however, should know better.
Similarly, anyone involved with a system would know it much better than a test user, yet as the user is an adult we still expect them to know better, although they don't stand a chance to know as much as us.
We learn from mistakes not from successes
As a teacher, I sometimes give a practical task to my students. The worse learning scenario is that the students perform the task without encountering a single issue. In the real world issues are common, so the more you are aware of possible mistakes, the less likely you are to make them.
This is also a key principle in design - "Fail fast, succeed sooner".
The remote control example
The previous point is something that is important to explain to users in the pre-test introduction. If time allows, I sometimes bring a complicated remote control with me and tell the user:
Imagine I try to understand design issues with this remote control, so I give it to you for 3 days, ask you to use it at home, then come back and report any issues you had.
Now imagine that you don't want to come across as stupid, so despite a multitude of issues you experience, you spend hours on end trying to figure things out and after 18 you succeed. Then you come back to me and tell me 'I had no issues - it's really easy to use, I can show you any feature on it'.
How would that serve real prospect users? How would it help us improve the product?
It is important to stress to participants that while we are delighted by success, it is difficulties that help us improve.
The user is not always right; users are
With all this in mind, it is important to remember that some participants' behaviour is nothing but extreme exception to the norm. This is why testing is ideally done with a good amount of subjects.
A user view that half the display should be a massive back button is unlikely to repeat with other participants.
So while the user is never 'wrong', a user is not always 'right'.