One possibly germane comment I can make is about the reliance on links to webpages or online resources in lieu of full and detailed citations. I've sometimes had folks come in behind my admittedly wordy source citations and reduce them to something like "1850 U.S. Census, Virginia, <link here>"; and some have not been happy when I reversed their abbreviations.
I've been dealing with the web since before the first graphical browser was introduced in 1993. And URLs change or vanish far more often than we may realize. We can look just a few years ago to Ancestry's acquisition of RootsWeb for an example; or the 2018 closing of WorldFamily.net due to the GDPR; or the wholesale changing of links at FamilySearch.org that occurred with their website revamp in 2011. All links are ephemeral and any of them could go away tomorrow; even the Internet Wayback Machine could lose funding and be forced to close.
Personally, I don't view documentation on WikiTree the same as I would for a research paper...but even on that paper I would need to use fully qualified citations that enable interested readers to locate, without an accompanying link, the exact source material I used. Instead, I view the WikiTree genealogy citations from the lens of an unknown 3rd cousin twice removed who is just becoming interested in investigating her family tree. If I use "1850 U.S. Census, Virginia, <link here>" can she locate the original information? Probably. Maybe even if the link has died. But she's going to have to do some digging.
If I've already done the work, why should I make her repeat it? Expanding that inadequate census citation to include everything that might be pertinent--enumeration district, dwelling number, street address if present, enumeration date (which can be vital to know but is often omitted in citations), the household members including all their warts of misspellings, incorrect ages, incorrect birth states, incorrect family relationships--means that the link to an original record becomes nothing but a nicety, just like it was a scant 20 years ago. There's no need for researcher after researcher to go slog through the very same handwritten entries over and over: I took a moment to record in detail what I found. They can verify with the original record at a glance if they choose, but the same wheel doesn't have to be repeatedly invented over and over. Life is short, and nobody is paying by the word to bind and print this stuff up into a book.
On one of my own websites I have a quotation from May 2013 by an unknown individual whose username at the Family Tree DNA Forums is "Frederator." He or she wrote:
"Oh, you wanted a short answer.
"Brevity, especially when it comes to something as operationally complex as the interpretation of autosomal DNA, comes at the expense of accuracy. I prefer accuracy."
WikiTree biographies can certainly fill up with GEDCOM junk; we've probably all run into profiles that were merged three or more times, each of which commingled more and more GEDCOM junk into an impressive morass. Thank you to the folks who try to clean that up for the rest of us. But a citation--in the correct place under Sources, not in the narrative biography--that provides more than the CMS minimum, that offers some transcription and possibly explanatory context...well, if it takes up several lines but presents solid information, my stance is that I'd rather have it than not, and that it's easy to scroll down on a webpage.
And yes, every relevant source discovered should be cited, IMHO. If an individual appears in six census enumerations plus a couple of census substitutes reconstituted from tax rolls, I think every single one of them needs to be cited. Genealogists reconstruct history piece by tiny piece, not just provide single citations to support BDM data.