2009/4/28 doc doc.wikipedia@ntlworld.com:
I disagree that the burden of proof is on the contributing author. The burden is on those wishing to delete something to achieve a consensus to delete. What level of "evidence" or "proof" will convince a consensus of wikipedians is up to the wikipedians participating. I suspect many people will be satisfied by different things.
The burden of proof has to be on the author. The person wishing to delete it would have to prove a negative, which is borderline impossible (in this case, anyway). How great the burden is is another question and, as you say, that depends on who is taking part in the discussion, but there is no choice about who the burden is on.