Another issue with online sources is the impermanence of URLs. Unlike the original concept, things change over time. Domain names are bought and sold or allowed to expire, site maintainers stop (other work or health), maintainers change and decide on a new schema. Even if there were a valid URL for every citation today (that is, every book, magazine, newspaper, etc was made available online), in a year half of them could be 404'd or simply contain different content.
There's a bit of a movement to encourage going back to Real Books(tm) for sourcing. But lots of wikipedians know about and lament linkrot, site death, and robots.txt hijacking (which can "disappear" an entire dead site from the Wayback Machine behind a domain squatter's temp page). So we soldier on with archive.org, WebCite, archive.is (new, and under indictment for bot spamming at the moment), and several other archivers to save our citations. Still, properly filled out references (author, title, publication, date, page) will always be verifiable even if its link rots, and one must take a trip to the library.
I understand, and I've noticed what seems to be a trend in WP articles of more "hard" sources. I think that's fantastic. My personal frustration as a would-be editor came from 3 types of edits that I would make that would frequently be reverted (years ago, and I haven't been back to edit since):
1. Grammar/typos - for whatever reason correcting "then" and "than" ticked some people off. "Could/would/should of", "intensive purposes", "loose/lose". These are common mistakes that happen with either careless writing or non-native speakers. The content can be good, but it feels like amateur hour and cleaning it up should be encouraged. Most of my changes of this sort were reverted within hours, even though I would include a comment in my commit explaining the changes.
2. Correcting based on the citations provided. Read an article, see something that seemed off, go to the citation. The citation states the exact opposite of the WP article (or some segment of the article), and I'd make changes based on that (usually it was careful editing so that it was a paraphrase or quote that dropped a "not" or something).
3. New content. I was new to WP editing, so I didn't know all the arcana (the process seems as esoteric to new editors as the old AD&D manuals seem to modern RPG players). I'd write something, post it, have the citations, realize I didn't put them in the day after. Everything was gone. I'd remake the changes with the citations, and it'd still be reverted (I honestly can't recall anymore which articles). I'm not doing research, there was no POV issues, I was just fleshing out content that was barebones. I'm the sort of person that could've been a good maintenance editor, filling in sparse articles or rectifying the information in others. Instead, the community seemed to actively reject my contributions. After a few starts like this, I saw no reason to persist.
For my own references, I have a git repo in my home directory, ~/references/, which has everything I consider important stored as PDFs that I "printed" via Firefox. It is the only way to make sure it is still around.
