If you are shooting gray card in camera to make test exposures, I recommend that you open the back of the camera in the darkroom and snip the sprockets to mark where the series end. Close camera and make another series etc. until the film is finished. Then when developing... Pull the reel, unwind and feel the first cut, and rip the strip off... continue developing a few minutes and pull the reel, feel for the next snip and rip. etc.
Not to discourage you from experimenting, but I don't think you need to "test" the effect of snip vs whole roll.
You are aware it's a variable that may exist - that's good enough to have awareness. Chances are you have many variables of the same magnitude - that are "out of control".
There IS a difference between tests and real runs. I deal with it, like I implied, by running tests at the tail end of regular developing runs. As statistics and process variability goes... none of my points are actually "on" the curve.
From my test "curve families" I draw a Time/Contrast Index chart. Each time I run a routine test, I'll add another point for the Time/Contrast Index chart. It's kind of like throwing darts. If it hits near where I aimed I'm happy. If it misses by a lot I start to look for mistakes.