Hopefully I understand your question correctly.
Just because you didn't show high price and highest specs together, doesn't mean people don't want both of those things (it also doesn't mean they wouldn't buy both of those things if you did put them together).
For example, we could do a study of mobile phones and we could prohibit Apple brand from showing with low prices. But Apple could still come out as the highest utility brand and the lowest prices could still come out as the highest utility prices.
Your prohibition affects what people saw but it doesn't restrict what they like.
Not that I think a restriction such as you describe is even a good thing: respondents wouldn't be confused by a low priced iPhone, they would just want to choose it.
Another thing to consider is that YOU might know enough about your market to know that your best specs can't be had for the lowest price, but the only way for your MODEL to learn that respondents value high specs and low prices is for it to see them make such choices. And the most efficient way for your model to learn that is if you use an efficient design that doesn't impose any more prohibitions than you absolutely have to have (i.e. prohibitions for logically impossible pairings, like a product that is both free and that has a $50 discount).