A propeller fan with a small number of blades (like 2 or 3) will impart regular pulsations to the air flow through it with each revolution of the propeller. Those pulsations will affect the accuracy of the measurements taken in the tunnel.
More blades raises the frequency of the pulsations, making them easier to filter out of the airflow upstream of the test section in the tunnel, but this will also introduce more "swirl" into the airflow, as the air tends to get dragged around in a spiral pattern with the rotating fan disc. This can seriously affect the airflow uniformity in the test section of the tunnel.
The usual filtration system is a series of closely-spaced wire mesh window screens just downstream of the fan, followed by a section of the tunnel which diverges into a very large cross-section, then converges again just before the test section.
Another way to minimize the effect of the fan on airflow dynamics through the test section is to place the fan on the exhaust end of the tunnel, leaving the window screens in the intake end of the tunnel.