This is my take on the basic question with a couple of examples from my own family research.
1)If the source referenced on wikitree just says 'family search' then in my opinion it's not worth the paper (screen?) it's written on.
2)If it's from one of the old pedigree or ancestral files or even their more recent tree, it may be valid but without further research you can't tell. Relationships may be based upon limited information or poorly understood documentation; great caution is needed. For example, my gg grandmother , a daughter of a very ordinary agricultural labourer has alternative fathers on these pedigree files.
Two of them say James was her father, two say Arthur. The ones that say James are wrong. They were probably based upon just one document, the 1841 census which doesn't state relationships. Charlotte was a child in 1841 and was with her brother at the home of James on the night of the census . Records of her baptism, the burial of her father and remarriage of her mother show that her father was dead and her mother living with a new family in 1841.Charlotte and her brother were staying or living with their uncle James. The LDS has been aware of this since the late 1960s and it was even written about as a teaching point for genealogy. Nevertheless, the error persists in half of these entries (and in consequence, other online trees)
3.If the source is to an index or a basic transcript of a source then it's got to be better than an unsourced tree but these indexes are a step removed from the original record. There are always going to be transcription errors; these especially occur when the transcriber hasn't much knowledge of the place.
Also, the strict format for filling in boxes when transcribing means that any extra information is not recorded We don't often know on marriage registers whether one of the pair was a widow(er),whether they came from that parish or were a sojourner or whatever.Actually we don't know whether the record was a marriage or a banns record so they can be quite deceptive. When I first started researching, such indexes were described as 'finding aids'
Indexes that miss extra information can also lead to gross errors.
"England Births and Christenings, 1538-1975", database, FamilySearch (https://familysearch.org/ark:/61903/1:1:J3WJ-C9Q : 21 September 2020), Joseph Cock, 1828
Seems like a good reference but Joseph Cock never existed.
The parish clerk made a mess of the register, he muddled two families so added them again, with a big arrow to point out the error. The transcriber has ignored the correction and just carried transcribing each entry . Fortunately, in this case, Family search has images of the original parish register. This shows the error; Joseph Satchell was baptised on the same day as Esther Cock. There was no Joseph Cock or indeed Esther Satchell register
This sort of thing is quite common in registers but FS indexes won't show it. Some transcription sites include a box for notes, this means that problems or extra information can be included (I wish FS would do this)
4)If the source on FS is an image of an actual document then it is a good source,hurrah!
But one such source may not be enough.Without other evidence, whether from Family Search or elsewhere it could lead you up someone else's tree. My gg granny does appear on the 1841 census at the home of James but he wasn't her father.