I agree about the evidence: it must be original, and it must be accessible (verifiable) by many - which means it has to be a scanned copy stored online of an original document with an electronic translation. Something like the census material on ancestry now.
I am generally aware of the limitations of DNA testing. I was (and am) trying to think of one way to put people of a common interest in one area (at least at first). I would include both biological and surname-based genealogy in this (at least I am interested in both). DNA alone of course doesn't provide the map. It is, at least currently, a way to check the written record for reasonableness and to suggest possibilities. As testing gets cheaper, 100+ sites can be sampled on a widespread basis, getting us well into where family records exist for many parts of the world.
On reviewers, I think you are comparing my "two" v. your "thousands". I would see thousands of reviewers also. I imagine they are not each reviewing the entire work. Ten years of work seems fine to me.
I'm not proposing a life re-creation effort. Getting the low-hanging fruit of census data and other semi-reliable records regarding an individual's birth date and location, parents (of both types), periodic locations, work, spouses, children, and death - a relable, maintained, updated framework that family researchers could use profitably for further work.
I'm of course unqualified to say how it should be done, but you guys aren't. It's hard for me to believe that an honest effort wouldn't be a substantial improvement over either of the alternatives available today.