It will. The calibration system doesn't look for the wand to be in all cameras, just in a certain minimum number. There are additional factors that can affect wand detection such as perceived angle and soforth, but for the most part if it's clearly visible in 20% of your cameras (minimum two), it may take a sample by default. That threshold is adjustable, but for the vast majority of solutions it should be appropriate.
So those cameras with less constant sampling shouldn't register lower on the "Excellent" "Good" "Poor" quality scale and/or show any anomalous readouts compared to the other cameras? I just want to be prepared for any odd readouts.