#1 is an internet marketing company. It has a website which verifiably exists. People considering doing business with them would find this useful. It should be kept.
#2 is an academic. That makes the article useful. OTOH, she's hasd death threats and wants the article down. I'm not sure what I'd do here.
#3 is a sportsman. Should be kept: even if he isn't well known, those who do know about him will likely find the article useful.
#4 is about Swiss people in Sri Lanka. If I was Swiss and was considering relocation to Sri Lanka, I might find this useful. Should be kept.
#5 is about "XMLmosaic", something I've never heard of, and has been speedily deleted.
A majority of these articles (at least 60%) are net-useful to humanity and should be kept. If I did a larger sample, I'm sure I'd get similar results.
About 100 articles a day are posted on AfD and most of these (75-80%) are deleted. What good is served by destroying tens of thousands of useful articles every year? None.
One good thing is that it keeps Wikipedia a tertiary source. I personally have a fairly inclusionist perspective, but my perspective is more or less: Wikipedia should cover anything on which any decent third-party source can be found (and it should point the reader to the source, so they can look it up themselves if they want).
There is of course stuff on which internet users can collect useful information, where no current sources exist to cite, but there's no reason Wikipedia has to be the only wiki on the internet. For example, Know Your Meme is an excellent project to document current internet memes, including via original sleuthing by its editors, who try to track down the history of memes and reconstruct their paths. It's not really a tertiary source summarizing the existing literature though; it's a sociological/historical research project. And it's great that it exists. Why does it have to exist on wikipedia.org, a project that has different goals, which don't include doing original research? Of course, memes should also be covered on Wikipedia, but only when Wikipedia can cite some existing published source documenting the information (a news article, a sociology conference paper, anything really).
I guess it seems perfectly fine to me that there's more than one project on the internet dedicated to collecting knowledge, with different goals; I don't see why Wikipedia has to be the union of all possible wikis. If anything, I would prefer more projects with different goals and approaches to exist, rather than everything being so centralized. That way I can go to Wikipedia if I want the tertiary-source take on internet memes, and I can go to Know Your Meme if I want the primary-source take on internet memes. Same with a Wikipedia article on a location versus a Wikitravel article on a location: both useful, but I don't see why they have to be merged into one project (or why that would be helpful).
Would you describe yourself as a hoarder? Applying your argument to personal belongings would appear to mean that unless it's actually completely unusable for anything it should be kept - even excrement is useful (very), as are old drinks cans (perhaps you could make a solar reflector), as are worn out clothes (make rags for cleaning), as are cellophane wrappings (could be heat welded and used as transparent insulation), ... sorry I'm digressing wildly.
Anyhow, the measure of inclusion on WP is not "isn't crap" but "is it notable" and whilst that gives a huge range to argue over it's is still clearly different to the question of "is it crap".
Take your sportsman - do you think that WP should be exhaustive? What is it about them that's notable and can be confirmed from other references?
Most of it isn't crap.
For example, I looked at the top 5 articles listed on 13th February -- http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion...
#1 is an internet marketing company. It has a website which verifiably exists. People considering doing business with them would find this useful. It should be kept.
#2 is an academic. That makes the article useful. OTOH, she's hasd death threats and wants the article down. I'm not sure what I'd do here.
#3 is a sportsman. Should be kept: even if he isn't well known, those who do know about him will likely find the article useful.
#4 is about Swiss people in Sri Lanka. If I was Swiss and was considering relocation to Sri Lanka, I might find this useful. Should be kept.
#5 is about "XMLmosaic", something I've never heard of, and has been speedily deleted.
A majority of these articles (at least 60%) are net-useful to humanity and should be kept. If I did a larger sample, I'm sure I'd get similar results.
About 100 articles a day are posted on AfD and most of these (75-80%) are deleted. What good is served by destroying tens of thousands of useful articles every year? None.