Not at this time. At one time we couldn't examine molecules either. I don't see why I need to tighten any definitions. If it exists then we should be able to study it if we are technologically advanced enough.
Without proper operational definitions, of course we cannot adequately study it, no matter how technologically advanced we are. This is why science has such a rich history of classifying things, and why it is so exciting to be able to use, say, genetic information rather than simple morphology to do so. If you and I both study NDE's, but due to lax definitional standards, you are using "retrospective accounts" and I am using "heart stoppage for 4 minutes or more" (other researchers may use different definitions than either of us), we will fail to find agreement about what a typical NDE looks like.
Seriously, I am dumfounded by the reluctance to use better experimental controls. Plain and simple: if there is a real effect, tighter controls will make it easier to see; if there is no real effect, tighter controls will show that too. If I were cynical, I could see this as researchers who know there is nothing there, trying to keep whatever source of funding they have. I am not that cynical, though, and so I am simply dumfounded. It's not that much harder to do the research well, than it is to do it poorly.