If one is forced by circumstance to work with a lab, as opposed to doing your own printing, the ideal situation (I think) is to establish a method of communication that works for both sides of the counter. You might, for example, start with a contact sheet, then have a straight "test print" done that would be the basis for discussing the objectives for the "final" print. The trick is probably providing enough guidance at each stage that the lab person knows what you're looking for. The obvious consequence, of course, is that the total investment in the final print reflects the work done by the lab.
I have to say that the approach that Bob Carnie describes is so far beyond the normal expectations that I can see why his competition may grumble about him. Certainly, that's part of his marketing, his way of differentiating his lab from the others, but I have to assume that his customers become very loyal as a result. Bravo, Bob.