Even if your customers are more likely to answer questions about what they want in the deal, it's still advantageous for you to soften the questions a bit. What you have to do, then, is look at your estimation of the list of items your customer is most likely to be concerned about, re-phrase the three questions above, and, as necessary, ask follow-up questions.
To answer the question: "Are these the right items?"
To answer the question: "What is most to least important?"
To answer the question: "How important are the top few items?"
Generally speaking, getting the answers to these questions enables you to determine how accurate your initial estimates were and, if necessary, adjust those estimates to reflect what you've learned. More specifically, the first question makes it possible for you to make sure that you and your customer are on the same page in terms of what items the customer is primarily concerned with. The second and third questions get those on the other side to recognize that some items are more important than others and, accordingly, which are critical and which might be traded away. Remember, trades come from the two sides valuing several items differently.
As with learning any new process, there are of course pitfalls to avoid. One of the most common pitfalls in preparing wish list validation questions is overcomplicating them. If the questions are too complicated, you get overcomplicated answers that yield little useful data. Alternatively, if you ask questions that are too broad, you wind up with too much general data. Ideally, ask simple, but specific questions, as these are the ones that are most likely to elicit the data you really need. Remember that ultimately what you want to know is what they want, what is most to least important to them, and how important the top few items are.
Best practice: always embed details of your estimation into your validation questions. That way, even if the other side doesn't answer them, at least you'll have shared some facts with them.