That is the question I had early on. With someone like Jim Dunn/Buddy Hull who lets be honest, are not going to contend for top half of the field, why push it and run a header that might be 40.005 degrees instead of playing it safe and running a 41 degree header angle especially with other teams being bopped earlier in the year. I do agree with their (Hull/Dunn) argument of the process for checking can be flawed and leave itself for inaccuracies however like Alan said earlier, if 20 cars were tested and 19 passed, it may not be the fault of the system used. However I feel like with all of the brilliant minds in the sports something more accurate could be developed. There is merit to both sides.For us, knowing the rules were drum tight, we knew our measurements might be +/- what NHRA's were. Early on we learned that the "smidge" needed to be pointed toward the safe side rather than up against the block wall enforced by NHRA. Served us well in the long run.
I believe one way to eliminate many of the variable's you mentioned is to measure using the front spindle and rear axle hub as the base line. That takes out the tire variation.I think they have to be measured on the car because of all the variables he and everyone else has mentioned. Tire pressures, driver weight, chassis differences, engine angle and on and on. A 40 degree header measured in a jig might measure 39 on one car and 41 on another depending how the motor is tilted or whatever the case is. I agree they need a solid tool and a flat place to measure each car (while theyre on the scale comes to mind) but it really can't be done off the car in my opinion.