smh, i call bullshit on the idea that speed tests are all over the place cuz of methodology. if you're doing it right, the results should be pretty consistent, no matter the niche. maybe some flukes or network variability, but that's not a methodology issue, that's just the nature of testing...