Mike Y is spot on. By nature, most courses converge to average. But, I will agree that there is no reason any more to build a poor course. The 50th percentile is probably moving up as most architects try to outdo each other.
I am not sure I agree with TD on the client relationship. They may or not care about what kind of golf course they get, but as long as they don't insert themselves into the process too deeply (which would probably be more likely with a client who cared, and was arrogant enough to think he knew more than the gca, and had just hired him to "draw up his ideas" it shouldn't make much different.
I recall being hired by a management guy, former tour pro, mostly for fee, in my early days. Halfway through shaping, he said he was buying a muni, but was getting a really, really good golf course. In other words, those who don't care can be sandbagged into a better course than they envisioned. And, I think most of us did that when possible, especially in the 90's when each new course had to be better than the last.
I have told the story before, but I spoke with one architect (not well liked here) whose prime design thought was to finish it in the three days he had allotted for the design. Another (not well known here, except to one poster) said it was too much work to avoid a few bad holes. A third kept making the same dull mistakes over and over. All the ASGCA meetings in the world, playing the best courses when he could, somehow, he could never translate good design over into his work, for reasons I couldn't understand.
So, architects attitude - or talent, lets face facts, not everyone who hangs out a shingle is equally talented - is probably such that it could affect end product quality. And, that talent might be partially deficient - either bad routing skills or no vision on feature designs.
For me, I do agree, and generally tried/try very hard to design a good golf course every time out.