My short answer is "Yes, but ..."
1) The "buts" include these:
* A well-sourced profile in an online family tree is a thing of beauty, The owner who has used original records, derivative records, and authored works to support their conclusions, meaning names, relationships, dates, places, etc. A profile like this can be believed after doing sufficient sanity checks on the information. Sanity checks might mean reviewing their sources, their media attachments, and their notes (if available).
* A poorly sourced profile in an online family tree is a thing to wonder about - if there are derivative sources and authored works, they should undergo a sanity check for the persons in question. In addition, the researcher should try to extend the research by searching for other records that might be available for the persons in question.
* An unsourced profile in an online family tree is a thing to be ignored.
* If an online tree had significant new information about a person, without source citation, I would contact the owner, if possible, to request more information, including their reasons fro drawing a specific conclusion.
* I think that source citations to authoritative records or documents is the key for believability. I don't even look at sources for other trees, but I do look at sources for actual vital records, military records, probate records, etc.
2) I trust original sources more than derivative sources, authored works, and indexed information. Even original records can be wrong in a detail. The other record levels have more of an opportunity to get a name or date wrong. Sometimes the original source is no longer available, and derivative sources must be used (e.g., probate clerk records, land recorder deeds, etc.).
3) The typical online family tree usually has several families that are well sourced, typically the researcher's birth family and grandparents' families, and perhaps earlier families. Then there are the families that may or may not be well-sourced, but the sources are typically vital or church record indexes, census records, Find A Grave memorials, etc. Many online trees have a number of families that are copied from other trees or from books and periodicals. This typical tree can have both believable and questionable conclusions because of the researcher's experience and interest, availability of sources, etc. It's a judgment issue, and a "sanity check" issue, I think.
4) My goal in my online family trees is to get my ancestral families as correct as possible, with source citations to original and derivative records, attached documents, and biographical and research notes. I also have a lot of profiles in my online trees whose information is based on indexed records, census records, and other online trees, mainly because I am collecting information about several of my key ancestral surnames - e.g., Seaver, Carringer, Auble, Buck, Vaux, etc. So my online trees are a mixed bag - some profiles are well-sourced, and others not so much.
5) I firmly believe that no family tree is ever finished. It is always in a state of correction and addition as more records become available and more research is performed.
I have over 46,000 Ancestry Hints to review at this time, of which over 33,000 are for records, not trees or photos. It takes hours to resolve 100 of these records - ignoring the ones that are from a poor derivative source, adding those from reliable sources to my database. Ancestry.com keeps adding Hints every day for my tree persons so it is a challenge to keep up. I typically ignore the tree Hints for persons that are not at the end of my ancestral lines.
The MyHeritage tree has 119,000 Record Matches, but two thirds of them are from other online family trees. I try to resolve these records occasionally, adding content and sources to my database.
I don't trust the online collaborative family trees like FamilySearch Family Tree, WikiTree and Geni when they extend back before 1600. I think that there are many errors in these trees before 1600, and there are some significant errors after 1600 in some of my colonial American lines too. It is a challenge to correct the pre-1600 lines.
6) What do my readers think? When do you use online family trees to get leads or clues?
Copyright (c) 2016, Randall J. Seaver