Question 3 is the one I would move to the top. A partner who can name the first small change they would ship, and roughly when, has already looked at your system. One who cannot is still selling a method.

I would add a ninth question: what exactly is priced, and what happens to the price when you learn something new? Modernisation work always uncovers rules hidden in stored procedures and in someone's memory, so the answer shows how risk will be shared. On client work we price the first release as one defined problem with a written definition of done, and treat whatever discovery turns up as a separate decision with its own price. Nobody is surprised by an invoice later.

Your point about asking for a reference from a project that went badly is excellent. The way a team describes its own failure tells you how it will behave in month four.

Which of the eight do buyers in the UAE tend to skip most often?