Tom,
I share your experiences. Usually, the super and/or management company comes into the project after the architect. They review everything from an operations perspective, including grass. As often as not, they recommend changes to their favorite selections for the area.
In some cases, this is influenced by local salesman, with whom the super has a preexisting relationship, or if a management company, a national buying agreement. I have lost out on several arguments (after all, the super is the one who has to grow it) with it working out both ways. At Colbert Hills, I favored A-4, and we used L93, which I am very happy with. We chose it as a medium maintenance grass, at the expense of a slight decrease in possible green speed and playability. The color is great, its fast enough,
There are so many turf choices today, it isn't easy to know which to use. Its usually not an all or nothng choice, but a matter of degrees - of color, speed, grain, disease resistance, etc. for greens, and most of the above and water useage, thatch, etc. for fairways. (The most typical debate for fairways is blue vs. bent)
On the other hand, I disagree with Stephen Biehl that a PhD is required! Give me someone who has selected grasses and grown grasses in the region on several courses versus someone who has studied it in the research fields!
Before I went to a researcher to select a grass, I would go down the street to see what a grizzled old superintendent was growing well.
A decade ago, golf architects chose greens mix.....then a bunch of soil scientists convinced everyone (including archies afraid of liability) that we weren't capable of that. Now, I see tissue testing, PhD in agronomy, etc. saying we don't know enough about grass to choose it? As mentioned above, it is more complicated, but it doesn't have to be that complicated.
Salesman and PhD's are great resources for learning about new grasses, but not always in the best position to make the call. Golf Architects should use both common sense, the supers input, and tried and true varieties for best results.
I think their is still some value in tryng new grasses in real life conditions, as opposed to research fields, especially in the cases where an architect is trying to do something different, like create links conditions. Of course, the Owner should be aware of the potential risks.