The page fails to distinguish the two fundamentally different usages of "dead" (as a term of strategy - as a rules term), to explain what "group" is, to consider stone versus string versus group, and to provide a clear, useful definition at all. Instead the page consists of half-truths.
Dead is a go term, but it is also natural language, and its usage reflects the informality and imprecision of natural language. Group is an even vaguer term. To say that the page consists of half truths is unduly critical, I believe.
Not that the page could not benefit from editing, but I do not think that we can talk about a single clear definition of a fuzzy term. Even the special definitions in official rules texts are not all that clear.
Nor should they be. There are examples of positions where it is plain that some stones are dead, but it is impossible to say which stones are dead.
If you like, dead can be defined for a) informal usage, b) formal strategic usage, c) as a rules term. Of course, official rulesets are the wrong source for (c); there are much more precise sources. While (a) may remain fuzzy, I have to disagree with you about (b). A clear definition there greatly helps players to understand what they are talking about. If you think that the impossible to which options should also be covered, there is scope for definitions (d), etc. that allow such.
Cher Robert,
There is only one definition given on the page. Why don't you provide, in addition, a formal, strategic definition? :-) As for the definitions in the official rules, links would probably be good enough, no?
Why? No time. I am being busy writing a new J1989 commentary, which is about the rules aspects of life and death definitions...