The dimensions shown are "nominal" values. I believe there was an interpretation from the committee that simply stated the values are nominal. That leaves it to the contractor to determine what tolerances should be applied. Some people elect to use the tolerances associated with the prequalified details found in clause 2, some use the groove tolerances listed in clause 4.

I use a 1/4-inch-thick bar to set the root opening. I cut the bevels with a bandsaw that has 1-degree graduations. Close enough for my work. It’s easy to get carried away with "tolerances" and "precision" when it really doesn't mean squat. Consider a standard set of fillet weld gages; do they really need to be "calibrated" when one considers the root opening can be up to 1/16-inch without correction or the fillet weld can be undersized by a specified amount depending on the size as long as it doesn't exceed some arbitrary percentage of the length. With that in mind, if a gage is worn and the fillet weld is 1/64-inch undersized, what difference does it actually make? And, where in the code are the tolerances of the gages listed? It “ain’t” in there, so who makes the call of what an acceptable tolerance is? The metrologist that calibrates the gages used by the tool and die maker?

Clause 5 allows the use of extension tabs. The width of the test plates are minimums, so I cut the plate width 7 to 8 inch instead of welding extension tabs to the ends of the grooves. I like the fact that the welders usually forced to make a start and stop somewhere in the groove when using SMAW. I like that.

With as many inspectors there are, you'll probably get the same number of answers and each will be a little different.

When I was welding for a living, I had a contractor that required me to torch cut my test plates and use a 6-inch grinder to prepare the bevel. Other job sites; the plates were precut on a bandsaw. The bottom line is the contractor is responsible to ensure the welder is properly qualified and the contractor is responsible for the welder’s work. They can make the test difficult or easier to suit their needed. 

"It ain't called the Farm Code for nothing." There's considerable latitude. Some contractors apply the latitude liberally, others, not so much. While I will take a “shot” at the code every once in a while, I think the approach taken by the code is fine. What difference is it going to make if the groove angle is +/- a couple of degrees or if the root opening is a little wider than the dimensions shown. The goal of the test is to see if the welder can deposit a sound weld. If the contractor cheats too much (and who is to say what too much is), he’s only going to make the test easier and may be putting welders to work that may be prone to producing substandard welds requiring a higher rate of rework. If the test is more stringent than necessary, good welders may “wash out” and the number of welders may be limited to a “choice few”.

I remember taking a test for a contractor that gave me what he called a Chinese Vertical. The plates were inclined at a 45° angle toward the welder, so it was actually considered an overhead by D1.1. A lot of welders washed out, but the welders that passed were pretty darn good. Did it meet “code”? Not in my book, but I was only the welder taking the test (and yes, I did pass.).

By the way, to answer one of your questions; I almost always use strong backs to limit angular distortion.

Al