When science doesn't match your personal observations, that doesn't make the science wrong. It also doesn't make your personal observations wrong - it usually (in my experience) means that you're measuring/observing different things (assuming you're reasonably intelligent, which I am doing). In this case, "cognitive development and well being" are complex concepts that probably do not map exactly (or maybe even grossly) between their definitions in the study and their experiential meanings in your head.
My personal observation is that "screen time" is not the problem, but lack of interaction _is_ - letting the kids be in front of a screen for two hours doesn't cause a problem, but letting that be an excuse not to spend time with them or interact with them (which is when it usually ends up happening in large quantities) does. Essentially - you need to control the other variables to understand what's really going on; excessive screen time is generally a symptom and not a cause.
My personal observation is that "screen time" is not the problem, but lack of interaction _is_ - letting the kids be in front of a screen for two hours doesn't cause a problem, but letting that be an excuse not to spend time with them or interact with them (which is when it usually ends up happening in large quantities) does. Essentially - you need to control the other variables to understand what's really going on; excessive screen time is generally a symptom and not a cause.