Average Constant?

No problem with using the average constant. Look at the bottom of the Bicycle Calibration Data Sheet. It clearly gives the OK.

Average constant has always been acceptable. Larger constant is usually easier (because it's usually the precal, and eliminates a postcal adjustment), and is recommended for a bit of extra safety, but has never been mandatory.

If possible, set a calibration course about halfway along the route. Measure to a convenient stopping point. Recalibrate, and measure to the finish from there.

We did this when we measured the Chicago Marathon early last month. The calibration course was close to the 14 mile mark. Worked like a charm. Even if you do decide to use the average constant, the numbers are closer together.

Plus, as always, if something goes wrong, you already have half the course in the books.

Another thing you should do is certify segments of the course. If the original measurer had done this, the entire course wouldn't need to be remeasured, just the segment(s) that contained the changes. (Note: the race Pete is measuring, the Hartford Marathon, is making minor changes due to traffic concerns)

BTW, I had a similar experience measuring the Mt. Rushmore Marathon & Half several years ago. When we'd calibrate at 6 or 7 a.m., the temps in the Black Mountains were in the 40s. By late morning, when we'd wrap up measuring, they were in the upper 70s or low 80s.