Hi, so I feel you. I don't often venture from my own tree now and generally only do so if it's a surname I have a good handle on, or in an area where I have a large personal tree only (Yorkshire, England). It's also why i haven't put my in-laws trees on.
However, when I am looking at something disconnected (as I've done quite a bit of work on graveyards near my Mum's house) is build a quick and dirty sideways large tree using the censuses - i have an ancestry subscription so in a couple of hours can build a Victorian tree out to a hundred or so profiles. I then import the gedcom and check for any potential matches. If there are none, I delete the gedcom and carry on widening. I focus on the second half of the nineteenth century because the records are easily accessible and it's basically add a couple of censuses for a family, marry off the easy children, add censues for the families of their spouses etc. I always focus on the easy obvious options (so would stop if I hit a Smith for example). I go sideways rather than backwards because the records are more readily available and I think you are more likely to connect into someone elses tree that way.
Once I've hit a potential connection I go back and work through just the profiles I need to create the connection to double check the quick and dirty part. At that point I tend to be using BMD indices to source. They are about the quickest way to very basically source a profile.
I can then go back to the slower, methodical work of adding in the profiles I want to work on. The light touch sourced profiles remain in my watchlist and I keep chipping away at those too (currently working on the earliest last edit date as a way of making sure I don't miss out any profiles).
Someone else has also mentioned FamilySearch, but I will often check for the profile there and hunt around the family a bit, often going backwards in time because I presume older ancestors are more likely to be someone's shared ancestors. As the other answerer has said, that still needs to be backed up with proper research, but identifying a potential route helps. I use other people's ancestry trees for this purpose too.
With both ancestry and familysearch it's often easy to spot what I would call "quality" profiles, where it looks like someone has spent more time doing their own research than just copying other people's trees. My "guess" is that they are the ones that are most likely to have had wiki-treer engagement and hence a slightly more likely route to finding a profile I can connect into.
Hope some of this helps.
