For research in Wireless Sensor Networks (WSNs) and the Internet of Things (IoT), while many protocols are either analysed mathematically or simulated to assess performance, it has become necessary that they are tested on hardware in real world environments. Algorithms are often validated in either (i) a densely connected network or (ii) a sparsely connected network. The majority of existing testbeds have implemented dense networks, making evaluation and validation of certain protocols, such as spatially-redundant source location privacy (SLP) protocols, challenging. We explore the use of transmission power to achieve network sparsity and present the results of such experiments performed on the FlockLab testbed. Based upon our experience in using FlockLab, we also identify additional aspects of testbed performance that need to be monitored to ensure reliable and reproducible results.