Honestly, I cannot think of a good reason to delete any article at all, unless it's obviously fraudulent, marketing-oriented, illegal, or obscene according to a widely accepted definition of obscenity. All of these standards can be applied fairly strictly, and with much less vagueness than notability.
- It's not like Wikipedia is short of disk space to store a few million extra text articles.
- The argument that it would be too difficult to maintain lots of extra articles is also weak, because not every article needs to be regularly edited, and more articles on niche topics might actually attract more editors.
- No, we won't end up with a page for every John Doe and his cat. That's just alarmism. Besides, if something like that ever becomes a problem, a better response would be a prohibition on self-promotion or some other clear guideline, rather than a vague requirement of notability.
- If these deletionists are just being OCD and wanting everything to be tidy and clean and under their editorial control, I would say that they need to take a break. In fact, it's possible that people with certain psychological traits self-select for Wikipedia editorship. But the kind of intolerance and self-centered narrow-mindedness that overzealous deletionists exhibit doesn't suit the spirit of a collaborative online project. Keep your OCD to your own home/office and away from public spaces, thank you very much.
Right now, I get the impression that it's too easy to flag something for deletion and too difficult to counter the deletionist argument, especially since the deletionists are so familiar with editorial procedures. This inequality needs to change. The burden of proof should be on people who want to remove information from the Web, not on those who want to keep it. Isn't that the same principle that we fought tooth and nail to uphold against the onslaught of SOPA, ACTA, etc?