Cadence will just be steady. It's possible that on longer runs (you basically just take readings for 2-4 runs, but alltogether) I could vary cadence... but more than anything it's a matter of sunken cost and would be the last run of the day most likely. (just to eat the final few minutes and, more importantly, to verify that I'm not moving/squirming and tainting the results)
Every run is going to cost roughly $25-$30 (small changes with a runner from tunnel to control room) and so it's easy to see how you just have to remove as many variables as you can and go with it.
Anything that takes any extra time starts eating into your $$$, so clothing changes (because I have to get out of the tunnel, change into a skinsuit and then get back in) cost more... doing something like testing 1 cadence vs. another cadence would basically mean testing 1 thing would cost $50-60. Testing wheels (because the tunnel operator is the only person who can touch the "rig" / change the wheel) costs more like triple anything else, which is why I'm reluctant to do more than one or possibly 2 changes.