Enlarge / Plume pods, each just big enough to plug into a wall outlet.

Part of the review process for Ars Technica’s first round of mesh testing was talking to the people and companies who build the devices. For established companies like Google or Netgear, you’re never sure who you’ll get—but for startups, who so far are casting long shadows in the Wi-Fi Mesh field, you’re likely to get some one-on-one time with the CEO or founder.

Plume was no exception, and I had a long talk with CEO Fahri Diner. Fahri made it clear from the start that Plume wasn’t about the typical Wi-Fi “ego yardstick”—i.e., setting up a single device super close to the AP in completely ideal conditions and trying to get the biggest number that you can. Real Wi-Fi workloads that real users put their equipment in charge of are an entirely different problem, he said, and that—not the e-peen measuring contest—was the problem Plume was designed to solve.

It’s easy to wave this kind of speech off as clever marketing BS. But I was more inclined to take it at face value than I might otherwise have been, since Plume had just done extremely well in my own testing. I try to make my testing as relevant to real-world conditions as possible—I test in a real house, with real-world distances, interior walls, exterior walls, and furniture in between nodes and client devices. Before using any artificial tools (like iPerf), I even test devices against real protocols (HTTP, SMB/CIFS, and so forth) to make sure they produce equivalent results.