RF in the Jungle: Effect of Environment Assumptions on Wireless Experiment Repeatability