Thank you for bringing this topic to light, if briefly.

You're absolutely right, and it's why I prefer to bring on employees who were contractors in one point in their life. If you can't grok, truly grok, that the business must be able to their jobs better... you're doing it wrong.

Sometimes you need to toss things in front of an end user as quickly as possible. Sometimes you see the long term effects and create a function to 'pass the salt'. You have to understand the long term business user. If you can't, you're going to drive me crazy when I follow you in.

Easy to work with doesn't mean bending over and taking it, however, which I don't believe you illustrated well in your very brief discussion. Easy to work with can mean explaining to the business the hurdles involved in reaching their goal, and helping them understand what the technical challenges will be in vernacular they will grasp. As a brief example, I've gotten half of the projects at my current position shot down after explaining why the difficulty of what they're asking for would take so much time with the tools currently in use.

To your specific point, I've learned to never say 'No' unless it was truly out of reach. I've learned to try to speak business' language in every one of my contracts, and explain the risks and cost. The decision in most cases must be left up to them if a $10/hour assistant is more profitable than what I, or another, can offer.